The Auchenipteridae and Pimelodidae of Suriname (Pisces, Nematognathi)

ienciennes.

THE AUCHENIPTERIDAE AND PIMELODIDAE
OF SURINAME
(PISCES, NEMATOGNATHI)

by

G. F. MEES

Rijksmuseum van Natuurlijke Histone, Leiden
With 41 text-figures and 15 plates

Contents

Introduction. 4

Zoogeography. 5

Bibliography. 7

Collecting localities. ®

Acknowledgements. 11

Auchenipteridae. 12

Auchenipterus .

Pseudauchenipterus . 2 3

Trachycorystes . 2 7

Auchenipterichthys .33

Par auchenipterus .3 8

Centromochlus .44

Tatia .55

Glanidium ..

Pseudotatia . io 5

Tocantinsia . io ^

Trachelyichthys . 111

Pimelodidae. 11 ^

Hemisorubim . I1 7

Br achyplaty stoma . 120

Pseudo platystoma . I2 7

Pimelodus . I 3°

Pimelodella . z 4 2

Rhamdia . z 5 2

Imparfinis .

Heptapterus . z 77

Pseudo pimelodus .

Microglanis . 221

List of Auchenipteridae and Pimelodidae known to occur in Suriname . . . • 2 37

New synonymy. 2 3^

Samenvatting. 2 59

References. 2 4°

Postscript. 2 47

Index. 2

4

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Introduction

This paper purports to give a review of the species of Auchenipteridae and
Pimelodidae known from Suriname. It is based on the large material collected
in recent years by M. Boeseman (1963/64), G. F. Mees (1965/66), and
H. Nijssen (1966/67), to which smaller collections from other sources have
been added.

Suriname, of course, is only a small part of tropical South America, prob¬
ably no more distinctive in its ichthyofauna than many other regions of a
similar size in this huge continent. Although I realized this from the beginning,
and was well aware that a serious study of the ichthyofauna of Suriname
would necessitate the examination of a lot of extra-limital material, I hoped
originally to confine this paper to descriptions and discussions of the species
actually known to occur in Suriname. Gradually, however, it became apparent
that such a restriction was difficult to adhere to, as the solutions to some of
the problems encountered in Suriname could be found only as far away as
Venezuela or south-eastern Brazil (if at all). Finally, so much information
on extra-limital material was collected, that it seemed a pity not to publish it.
Whereas, admittedly, the inclusion of all this material has made the paper
less conveniently arranged, I believe that these additions have increased its
value. To reduce the danger that the species actually known from Suriname
are drowned in the discussions of extra-limital species, a list with their names
is appended.

Where so much revisional work had to be done anyway, one could justifiably
ask why I have not taken the logical next step and attempted complete revi¬
sions of the families Auchenipteridae and Pimelodidae. The reply to this
question is in the first place that this would be a task of a magnitude very
probably beyond my powers: of some of the genera ( Pimelodella , Rhamdia)
over fifty species have been described (many of them only known from their
type-material, which it would often be difficult to obtain for examination),
and in the second place that even if it was possible to bring all extant material
together, this would probably not suffice for a proper understanding of cer¬
tain genera (for example Pitnelodus). Even in the genera which I have
studied, it was often impossible to obtain the necessary material, either because
not enough specimens exist, or because of regulations which forbade loaning
type-specimens, or, especially in the case of certain South American institu¬
tions, because my letters of enquiry remained unanswered.

If this paper contributes only a little to the understanding of the groups
treated, it will have fulfilled its purpose.

In the keys and generic diagnoses I have tried to be comprehensive, so that
the characters are also valid for species not treated in this paper. In many

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

5

instances, however, I was unable to attain this (for example in the key to the
Pimelodidae, which would have become much more complicated, had all extra-
limital variation been considered). Keys and descriptions published by others,
for example in the paper by Schultz (1944a), suffer from the same short¬
comings, but nevertheless are widely used.

The material collected during the first three months of 1971 in the Nickerie
District, north-western Suriname, by M. Boeseman and D. C. Geijskes (cf.
Geijskes, 1973) has become available too late for comprehensive inclusion in
this report, but I have looked part of it through, and take this opportunity
to mention the more interesting species. The collection contains a series of
Tatia creutzbergi from the Nickerie Rivier, and one from a tributary of the
Maratakka (locality indicated in fig. 19); two samples of Pimelodella mac-
turki from the Bigi Birri Kreek, a tributary of the Maratakka, a very wel¬
come confirmation of the occurrence of this species in Suriname,, and, perhaps
most interesting of all, three specimens of Pseudopimelodus raninus villosus
from the Winanna Kreek, a tributary of the Kaboeri Kreek, which in its turn
flows into the Corantijn. This record extends the range of the subspecies from
the Demerara River eastwards to the Corantijn, and into Suriname.

Except for the introductory chapters, this paper was completed around the
middle of 1971, and no attempt has been made to add references to literature
and material received since then, with a few exceptions, most of which have
been added as footnotes.

Zoogeography

A zoogeography of the freshwater fishes of Suriname must be based on its
whole ichthyofauna. Nevertheless, a short discussion of the two families here
dealt with, will contribute to an understanding of the whole.

For an analysis I have divided the 30 species into three groups: species which
on present evidence are endemic to one river system; species confined to the
Guianas, and species also occurring in the Amazon basin.

The first group contains: Tatia concolor, Glanidium leopardus, Heptapterus
tapanahoniensis , Heptapterus surinamensis.

The second group contains: Tatia brunnea, Pimelodella macturki, Heptap¬
terus brevior, Heptapterus bleekeri, Pseudo pimelodus nigricauda, Microglanis
secundus.

The third group contains all other species; thus of the 30 species, 20 are
known also to occur in Amazonia (the position of P . blochii needs clarifica¬
tion, it may have to be excluded). Probably the species Pseudo pimelodus nigri¬
cauda and Microglanis secundus , the first of which has been found through¬
out Suriname, the second in Suriname as well as in British Guiana, are much

6

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

more widely distributed, as it is unlikely that they would stop at the artificial
political boundaries. SeveraJl species which were hitherto believed to be con¬
fined to the Guianas, are in this paper recorded for the first time from Ama¬
zonia (Tafia creutzbergi, Microglanis poecilus ) x ), and it is very likely that
further ichthyological investigations in northern Brazil would increase this
number. In other words, the first two lists (endemics of one river system, and
endemics of the Guianas) are probably much flattered. It looks as if only
Heptapterus has some species which are genuinely endemic to a single river
system: H. surinamensis being confined to the Suriname Rivier basin, H .
tapanahoniensis to tributaries of the Marowijne.

The close affinity between the Guianas and the Amazon basin has been
stressed by several workers, notably Eigenmann (1912: 73), and in recent
years by Myers.

The fact that species inhabiting the lower courses of the great rivers are
widely distributed, will surprise nobody, but it is interesting to Observe that
small species which normally appear to be confined to rapids, may also have
a very wide distribution. Imparfinis minutus is an example of this.

It is remarkable that of some other groups of nematognathous fishes a high
degree of local endemism appears to exist in Suriname, see the works of Nijs-
sen & Isbriicker (1967) and Nijssen (1970) on Corydoras (ism. Callichthy-
dae) and of Boeseman (1968, 1969) on Hypostomus (fam. Loricariidae).
Hypostomus constitutes an extreme example of this as not only each river
system appears to have its own species, but there are even differences between
the tributaries of one river (viz., Kabalebo and Sipaliwini, tributaries of the
Corantijn). This difference between Auchenipteridae and Pimelodidae on one
side, Corydoras and Hypostomus on the other, can probably be ascribed to two
factors. One is an ecological one: most species of Hypostomus are confined
to rapids, which would restrict their chances of colonising other river systems
(when other species might make use of temporary flooded areas, estuaries,
etc.), and the species of Corydoras are small bottom-fishes, presumably some¬
what sedentary in habits. In this connection it may be significant that Heptap¬
terus , which in its preference for rapids shows an ecological resemblance to
Hypostomus, is the only pimelodid genus in which endemism is found in Suri¬
name. The second factor is morphological: both Hypostomus and Corydoras
are armoured with heavy bony plates, and usually have black spots or bands,
and other complicated pigment-patterns. Small differences in proportions and
in pigmentation would be easily noticed and measured. In the soft-bodied

1 ) Admittedly Microglanis poecilus had heen recorded from the “?Amazon basin” by
Gomes ( 1946 : 15 ), but the specimens are M. parahybae, and probably from south-eastern
Brazil.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

7

Auchenipteridae and Pimelodidae, on the other hand, differences in propor¬
tions would have to be much greater to be measurable, whereas at least in
some genera ( Pimelodella , Heptapterus ) the pigmentation is so uniform, that
geographical variation of a kind useful for systematic distinction, is less
likely to occur.

In this connection it is also necessary to state that in this paper I have been
more concerned with arriving at a synthesis: placing together what belongs
together, than with analytical work. Moreover, whereas of some widely
distributed species large series from Suriname were available, usually few
specimens, from scattered localities in other parts of their ranges, could be
examined, so that small but constant differences in measurements and pro¬
portions between various populations would for that reason have remained
undetected. It is very likely, therefore, that future analytical work will show
that certain widely-ranging species, for example Parauchenipterus galeatus,
Rhamdia quelen, etc., can be divided into subspecies. In a few instances, in
Pseudopimelodus zungaro and Pseudopimelodus raninus , I have already
attempted (perhaps not too successfully) to express the geographical variation
which undoubtedly exists, with the application of trinomials.

Bibliography

Bibliographies of South American Nematognathi have been published
repeatedly, notably by Eigenmann & Eigenmann (1890), and more recently
by Fowler (1951). Such bibliographies are tremendously useful to anybody
wanting to orient himself quickly in the literature of a particular species. In
view of the comparatively recent date of Fowler’s compilations one might,
however, wonder whether it was necessary for me to list again all those refer¬
ences. Apart from the fact that Fowler’s bibliography is confined to the Ama¬
zon basin, and does not include the Guianas, my excuse is that Fowler’s work
was largely uncritical: in most cases he has just accepted published records
as being based on correct identification. I, on the other hand, have re-examined
a large proportion of the material on which records of the species here dealt
with were based, and have attempted to assign them correctly. How much
difference that makes will be seen in the cases of the two Centromochlus-
species, Tatia intermedia, Glanidium melanopterum, etc., where a majority of
references proved erroneous.

Of course, not all references could be verified by an examination of the
material on which they were based; when I believed such references to be
acceptable, I have included them without comment. When I did not trust
them, a discussion is given. Undoubtedly, there remain references incorrectly

8

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

placed, just as several of my systematic conclusions will probably prove to be
mistaken.

It has not been possible to make complete bibliographies: for example the
comprehensive work of Ringuelet, Aramburu & de Aramburu (1967) con¬
tains references to a number of South American publications which have not
been available to me.

Collecting localities

This is not the place for an extensive treatment of the topography of Suri¬
name, but for a good understanding of the zoogeography and composition of
its ichthyofauna a few facts have to be mentioned. Suriname, part of Guianas,
has an area of 161900 km 2 (including disputed territories: 13583 km 2 in the
South-West and 3441 km 2 in the South-East) ; like all tropical countries it is
well watered. All rivers flow in a northerly direction into the Atlantic Ocean.
As the southern (Brazilian) border follows the watershed between the nor¬
thern (Atlantic) and Amazon drainages, no part of the Amazon basin is
included in the country, and there is no direct connection between the rivers
from Suriname and those of the Amazon basin. This situation differs from
that of Guyana (British Guiana) and Venezuela: in the south-western part
of British Guiana the Ireng and Tacutu rivers, tributaries of the Rio Branco,
belong to the Amazon basin, and it is of course generally known that the upper
course of the Orinoco in Venezuela is, through the Casiquiare, broadly con¬
nected with the Rio Negro. Although a direct connection with the Amazon
does not exist in Suriname, it is nevertheless likely that fishes are able to
cross the border. In 1972 I spent the middle of the wet season (May-July) at
Sipaliwini near the southern border, and found large areas of forest and
savanna flooded, with various species of fish plentiful in such temporarily
flooded areas. From the report of the boundary commission (van Lynden,
1939) it is apparent that the watershed between the Sipaliwini and the Rio
Paru del Oeste (which forms the boundary between Suriname and Brazil)
is scarcely above the level of the surrounding swamps, and in the wet season
would not form a veiy formidable barrier to aquatic animals.

Although we have tried to make our collections comprehensive, and material
from all major river systems has been obtained (for general orientation, see
fig 1), our ichthyological studies took place in the framework of the “Broko-
pondo project”, which aimed at investigating the changes in flora, fauna, and
physical properties of habitat, resulting from the damming of the Suriname
Rivier (for general information on the project, see van der Heide, 1966, and
Westermann, 1971). We were stationed at Brokopondo on the Suriname
Rivier, about 10 km downstream from the dam, and the intensity of our

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

9

activities in the Suriname Rivier basin (including the lake, as it gradually
filled) was much greater than anywhere else in Suriname. Fig. 2 shows the
Suriname Rivier, on a larger scale than in fig. 1, with many localities which
could not be crowded into the general map.

As our investigations took place before and during the filling of the lake,
it has been necessary to indicate on the map several localities now submerged.

The majority of the localities of the old material re-examined by me can
be found on the maps published by Agassiz & Agassiz (1869) and Eigenmann
& Eigenmann (1890), and in Eigenmann (1911).

As this is not a treatise on geography, I have not attempted to attain uni¬
formity in spelling of geographical names, and have usually copied the spelling

10

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

II

as it appeared on labels or in publications. Also, I have preferred to use the
former name of British Guiana for what is now the State of Guyana, as the
name Guyana could be misleading, being applicable to the whale area between
the deltas of Orinoco and Amazon.

Acknowledgements

The institutions from which material was borrowed, with the abbreviations
used for them, are listed below. To all the persons named I am indebted for
the generous way in which they have placed material at my disposal.

AMNH American Museum of Natural History, New York (Dr. D. E.
Rosen, Mr. J. N. Baskin).

ANSP Academy of Natural Sciences, Philadelphia (Dr. J. E. Bohlke).
BM British Museum (Natural History), London (Dr. Rosemary H.

McConnell, Mr. G. Howes).

CAS California Academy of Sciences, San Francisco (Dr. W. N.

Eschmeyer, Dr. W. I. Follett).

FM Field Museum of Natural History, Chicago (Dr. L. P. Woods,

Mr. R. Schoknecht).

KBIN Koninklijk Belgisch Instituut voor Natuurwetenschappen, Brus¬
sel (Dr. J. P. Gosse).

MACN Museo Argentino de Ciencias Naturales “B. Rivadavia”, Buenos
Aires (Dr. H. P. Castello).

MCG Museo Civico di Storia Naturale “Giacomo Doria”, Genova
(Prof. Dr. E. Tortonese).

MCZ Museum of Comparative Zoology, Cambridge, Mass. (Miss
M. M. Dick).

MLP Museo, Universidad Nacional de la Plata, La Plata (Dr. R. H.
Aramburu).

MNP Museum National d’Histoire Naturelle, Paris (Mme M. L. Bau-
chot).

MRJ Museu Nacional, Rio de Janeiro (Dr. H. Sick).

MTD Staatliches Museum fur Tierkunde, Dresden (Mr. W. Hebig).
MV Naturhistorisches Museum (Museum Vinogradensis), Wien (Dr.

P. Kahsbauer).

MZSP Museu de Zoologia da Universidade de Sao Paulo, Sao Paulo
(Dr. H. A. Britski).

NMB Naturhistorisches Museum, Basel (Dr. E. Sutter).

RMNH Rijksmuseum van Natuurlijke Historie, Leiden (Dr. M. Boese-
man).

12

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

SMF Natur-Museum und Forschungs-Institut, Senckenbergische Na-
turforschende Gesellschaft, Frankfurt a. M. (Dr. W. Klause-
witz, Dr. F. Rossel).

SU Stanford University, Stanford, California (Prof. G. S. Myers,

Dr. W. C. Freihofer).

UMMZ University of Michigan Museum of Zoology, Ann Arbor (Mr.
G. R. Smith).

USNM United States National Museum, Washington, D.C. (Dr. R. H.
Gibbs, Dr. V. C. Springer).

UZMK Universitetets Zoologiske Museum, Kobenhavn (Dr. J. Nielsen).
ZMA Zoologisch Museum, Amsterdam (Dr. H. Nijssen).

ZMR Zoblogisches Museum der Humboldt-Universitat, Berlin (Prof.
Dr. K. Deckert).

Dr. J. Gery (St. Cyprien, France) has to be thanked for donating a few
interesting specimens, amongst which a Tatia from the Rio Paru de Oeste,
zoogeographically of great interest, and Professor M. P. de Godoy (Estagao
Experimental de Biologia e Piscicultura, Pirassununga, Est. de Sao Paulo,
Brazil) for information on certain species of fish occurring in the Mogi
Guassu.

The figures nos. 2, 7, 9, 11-14, 16*26, 28, 30, 35, 36 and 41, and the plates
nos. 6, 8, 10 and 11, are from the hand of our artist Mr. W. C. G. Gertenaar.
For the remaining figures I am responsible. The frontispiece was received from
Mme M. L. Bauchot, and the other photographs (plates 1, 2, 7, 9 and 12) were
taken by Mr. Chr. Hoorn. To Mr. J. Simons of the Zoologisch Laboratorium,
Leiden, I am indebted for X-ray photographs of certain specimens.

Auchenipteridae

The Auchenipteridae form a possibly heterogeneous assemblage of naked
catfishes. Teeth in bands in both jaws, none on vomer and palatines, except
in Asterophysus and occasionally in large specimens of Tatia intermedia ;
nuchal bones exposed or covered with skin only, reaching to the dorsal spine
and partly embracing it with processes which in the species descriptions will
be referred to as “horns”; three pairs of barbels, well-developed but not large,
the largest pair, that of the maxillary barbels, not reaching much beyond middle
of body (usually shorter), an exception is Tetranematichthys quadrifilis
which has only a single pair of very small mandibular barbels; no free eye-rim
although the eyes are often large; gill-openings restricted, reaching to just in
front of cleithrum, membranes broadly attached to the isthmus (except in two

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

13

little-known monotypic genera, Wertheimeria and Taunayia , which have the
gill-openings extended forwards along the isthmus; these genera require
further study); pectoral pore present in all species studied, but often very
inconspicuous l ); strong pectoral and dorsal spines; long or short anal fin,
and a usually small adipose fin, which is entirely lacking in some genera.
Most genera show a certain degree of sexual dimorphism, mainly in the
'shape of the anal fin.

Some authors, beginning with A. de Miranda Ribeiro (1911: 352, 394)
have divided the Auchenipteridae as here understood in two families, the
Auchenipteridae (s.s.) and the Trachycorystidae. Others (Gosline, 1945) on
the other hand, have treated them as a subfamily of the Doradidae. The latest
reviser is P. de Miranda Ribeiro (1968b) who has split the Auchenipteridae
in no less than four families: Auchenipteridae (s.s.), Trachycoristidae, Cen-
tromochlidae, and Asterophysidae.

My personal opinion is that with the possible exception of some small genera
which I have not yet studied ( Asterophysus, Wertheimeria, Taunayia, and
perhaps Tetranematichthys) , the Auchenipteridae form a natural group. Also
in view of the comparatively small size of the family as a whole (no more
than 60 species in 19 genera), the differences between P. de Miranda Ribei-
ro’s four families would be adequately expressed by retaining them as tribes,
or subfamilies at most. Admittedly, it is likely that many more species remain
to ibe described but even if its number of spedes were doubled, the group
would not become too unwieldy to be retained as a single family. As I have
not yet studied the Doradidae, I postpone giving an opinion on the affinities
between this family and the Auchenipteridae, but wish to draw attention to
Liosomadoras, discussed below, which appears difficult to place. Indeed it
was the intermediate position of this genus which induced Gosline (1942) to
combine the Doradidae and the Auchenipteridae.

Rossel (1962) has suggested that the species Liosomadoras morrowi Fowler
(1940a: 226-228), described in the Doradidae, belongs to the Auchenipteridae,
and that Liosomadoras may be a synonym of Centromochlus (i.e. Tatia). I
have not examined the species, but, whereas Rossel may possibly be right in

1 ) Steindachner ( 1876 b) and Eigenmann & Eigenmann ( 1890 ) have described several
species of Auchenipteridae as lacking a porus pectoralis. I have not studied all these
species but in those which I did examine, a pore was invariably present; as this pore is,
however, often very small and difficult to find, especially in species in which it is placed
just above the pectoral base, I venture to suggest that the above-mentioned authors over¬
looked it in the other species. The porus pectoralis may be of considerable systematic and
phylogenetic significance, and certainly deserves to be carefully studied, but is too easily
overlooked to be of much use in routine identification. In many descriptions it has not
even been mentioned.

14

ZOOLOGISCHE VERHANDELINGEN 17,2 (1974)

the familial allocation, Fowler’s description and figures show that the species
is certainly sufficiently distinct from both Tatia and Centromochlus to be
retained in a separate genus. From 'both these genera, Liosomadoras differs
by its short body with very broad flat head, small eyes, superior mouth, large
adipose fin, large anal fin with 13 rays, and very long spiny humeral process;
from Tatia moreover in the long dorsal and pectoral spines and a well-
developed caudal peduncle, from Centromochlus in having dorsal and parti¬
cularly pectoral spines provided on both sides with strong teeth.

The fish described as Trachymochlus cupido by Hoedeman (1961), and
referred by him to the Auchenipteridae, does certainly not belong to this
family. Apart from the characters already given by Hoedeman: the presence
of nasal baihels and of palatinal teeth, it differs from the Auchempteridae
by: free orbital rim, gill-membranes free from isthmus, large adipose fin,
occipital process without horns, reaching to the tip of the well-developed
triangular predorsal plate, etc. On external morphological characters the
species would appear to be well-placed in the Bagridae, a family not known
from South America; it is possible that an anatomical examination would
reveal its true affinities, but this must await the collecting of additional
material. For the present it must go as genus incertae sedis. As, however, the
species was described in the Auchenipteridae, mention of it had to be made
here, to justify its omission *).

From Suriname, nine species of Auchenipteridae are known, divided over
five genera: AuchenipterUrS (one species), Pseudauchenipterus (one species),
Parauchenipterus (one species), Tatia (five species) and Glanidium (one
species). To arrive at a proper understanding of these genera I have also had
to study many extra-limital species and genera, whereas it was thought that
a simple key to the known genera would be considered useful.

1 ) The Cupido Kreek is a tributary of the lower Maratakka and the type-locality of
Trachymochlus cupido is supposed to be the mouth of the Cupido Kreek, at the place
where it joins the Maratakka. In the first half of 1971 , Boeseman and Geijskes collected
fish at several localities in the Maratakka basin, although not in the Cupido Kreek itself,
and found no trace of Trachymochlus cupido. Dr. Nijssen has informed me that, as far
as he can see, the provenance of the specimen is above suspicion: the collector, J. van
der Kamp, is known to have visited the Cupido Kreek on the date indicated, and has
presented two specimens of fish from there to the Zoologisch Museum. On the other

hand, Nijssen observed that in those days, after the specimens had been received,

“anything might have happened”.

In the circumstances, and in view of the improbability that a fish unlike any South
American family, but obviously close to a family known from the Old World, would
have remained undiscovered except for this single specimen, I believe that I can be

more positive than as expressed in the text above, and that on present evidence it is

justified to remove Trachymochlus cupido from the South American fauna.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

15

Key to the genera of Auchenipteridae *)

ia — Gill-openings restricted, reaching to just in front of cleithrum 2
b — Gill-openings continued well forward along the isthmus . 16

2a — An adipose fin present (sometimes small).3

b — No adipose fin.13

3a — Ventrals with 9-13 rays, anal with 57-59 rays . Pseudepapterus

b — Ventrals with 12-15 ra y s > anal with ( 3 2 ~) 37"49 ra y s Auchenipterus

c — Ventrals with 6-10 rays, anal with less than 50 rays . . 4

4a — Ventrals with 10 rays, anal with ca. 13 rays . . Asterophysus

b — Ventrals with 9 or 10 rays, anal with 20-26 rays ... 5

c — Ventrals with less than 9 rays.7

5a — Caudal fin obliquely truncate or shallowly forked ... 6

b — Caudal fin deeply forked: aberrant specimen of Pseudauchenipterus
6a — Head strongly depressed, with bones exposed and granular; fontanel
small, roundish-oval; eyes small, more than six times in head

. Trachycorystes

b — Head only anteriorly depressed, its posterior part almost as high as
wide; fontanel elongate, over three times as long as wide; eyes large,
less than four times in head .... Auchenipterichthys
7a — Ventrals with 8 rays, anal with ca. 20 (17-25) rays .

. Pseudauchenipterus

b — Ventrals with less than 8 rays.8

8a — Ventrals with 7 rays.9

b — Ventrals with 6 rays.10

9a — Anal with 12 rays, head depressed, its depth twice in its width

. T ocantinsia

b — Anal with 15-16 rays, head not depressed, its width measured between
the deithra only a little more than its depth at the same place

. Pseudotatia

10a — Anal with 17-41 rays.11

b — Anal with 7-13 rays.12

na — Maxillary barbels long, reaching to base of anal fin, inner mandibular

barbels reaching to pectoral base, anal with 18-21 rays, caudal forked
with pointed lobes, body moderately slender . . Entomocorus

b — Maxillary barbels at most to below dorsal fin, inner mandibular
baihels not nearly reaching to pectoral base, anal with 17-41 (usually
20-30) rays, caudal fin obliquely truncate or obliquely rounded, some-

1 ) Tetranematichthys , easily recognised by having only one pair of mental barbels, is
regarded as a generic synonym of Par auchenipterus by P. de Miranda Ribeiro ( 1968 c).

l6 ZOOLOGISCHE VERHANDELINGEN 13 2 (1974)

times shallowly emarginate, but never forked, body short and heavy

. Parauchenipterus

12a — Pectorals with one strong spine and 7-12 rays, predorsal length 2.7 to

3.25 times in standard length. Centromochlus

b — Pectorals with one spine and 4 or 5 rays, predorsal length 3.25 to

3.7 times in standard length. Tatia

c —• Pectorals with one spine and 5 or 6 rays; predorsal length 3.2 to
3.7 times in standard length; very similar to Tatia , for differences

see generic diagnosis.. Glanidium

13a — Ventrals with 6-10 rays, anal with less than 40 rays ... 14

b — Ventrals with 14-16 rays, anal with over 50 rays ... 15

14a —• Ventrals with 6 rays, anal with ca. 32 rays, eyes small, 6 times

in head. Trachelyopterus

b — Ventrals with 10 rays, anal with 35 rays, eyes large, less than 4

times in head. Trachelyichthys

15a — Ventrals with 15 or 16 rays, anal with ca. 53 rays, dorsal with a spine
and four rays, teeth present .... Trachelyopterichthys

b —- Ventrals with 14 rays, anal with ca. 61 rays, dorsal with a spine and

three rays, no teeth. Epapterus

16a — Anal with n rays. Taunayia

b — Anal with 15-16 rays. Wertheimeria

Genus Auchenipterus Valenciennes

Auchenipterus Valenciennes, 1840, in Cuvier & Valenciennes, Hist Nat. Poiss.
(4 0 ed.), 15: 154 — type by original designation Hypophthalmus nuchalis Spix.

Euanemus Muller, 1842, Ber. Verh. Akad. Wiss. Berlin: 203 — type by original
designation and monotypy Euanemus colymbetes Muller = Auchenipterus nuchalis
(Spix).

Ceratocheilus A. de Miranda Ribeiro, 1918, Rev. Mus. Paul., 10: 644 — type by mono¬
typy Ceratocheilus osteomystax A. de Miranda Ribeiro = Auchenipterus nuchalis
(Spix), $.

Osteomystax Whitley, 1940, Austr. Nat., 10: 242 — nomen novum for Ceratocheilus
Ribeiro, preoccupied by Ceratocheilus Wesche, 1910.

Generic diagnosis. — Slender, body compressed, much deeper than wide,
except for the head which is almost round in circumference; eyes large, their
position lateral, but with a distinct downward direction; pectoral pore over
middle of pectoral base; dorsal and pectoral spines thin and pungent; ventrals
with 12-15 rays, broadly attached to the body; anal fin long, with (32-)37-49
rays; adipose fin very small; lateral line complete, consisting of a number of
loops, and giving off branches. At least one species shows sexual dimorphism
in shape of maxillary barbels and anal fin, and length of dorsal spine.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

17

Colour plain brownish or grey above, unpigmented below, with or without
a silvery lateral band.

Distribution. — Eastern South America, from British Guiana to Uruguay,
including the Amazon basin.

Discussion. — Not counting 'six names which I regard as synonyms, and
which will be dealt with in the discussion of A. nuchalis, the genus contains
the following five species: A . nuchalis (Spix, 1829); A. brachyurus (Cope,
1878); A . fordicei Eigenmann & Eigenmann, 1888; A. nigripinnis (Bou-
lenger, 1895); A. paysanduanus Devincenzi, 1933. It is doubtful whether all
these species are valid: whereas the descriptions of A. nigripinnis and A.
paysanduanus look convincing, A. brachyurus and A. fordicei, each known
from a single specimen collected over a century ago, should be re-studiod in
the light of present knowledge x ). P. de Miranda Ribeiro (1968a) has sug¬
gested that A. brachyurus, A . fordicei, A . nigripinnis and A. paysanduanus
are all synonyms of A. nuchalis: if this is confirmed, the genus Auchenip-
terus would be monospecific. A. paysanduanus was described as having only
32 or 33 and rays (Devincenzi, 1933), which is below the normal variation of
A. nuchalis, and suggests its validity.

Auchenipterus nuchalis (Spix) (fig. 3)

Hypophthalmus nuchalis Spix, 1829, Gen. Sp. Pise. Bras.: 17, pi. XVII — Brasilia
aequatoriali.

Auchenipterus dentatus Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.), 15: 156 •— Cayenne.

Euanemus colymbetes Muller, 1842, Ber. Verh. Akad. Wiss. Berlin: 203 — Surinam.

Auchenipterus demerarae Eigenmann, 1912, Mem. Carnegie Mus., 5: 202, pi. 21
fig. 1 — Wismar, Demerara, British Guiana.

Auchenipterus brevior Eigenmann, 1912, Mem. Carnegie Mus., 5: 202 — Tumatumari,
Lower Potaro River, British Guiana.

Auchenipterus ambyiacus Fowler, 1915, Proc. Acad. Nat. Sci. Philad., 67: 222, fig. —
Ambyiacu River, Ecuador.

Ceratocheilus osteomystax A. de Miranda Ribeiro, 1918, Rev. Mus. Paul., 10: 644 —
Santa Rita das Antas, Rio Vermelho, Goyaz.

Auchenipterus nuchalis Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat.
Poiss. (4 0 ed.), 15: 155 (Bresil); Filippi, 1853, Rev. Mag. Zool., (2) 5: 167 (no
locality, ex Spix) ; Castelnau, 1855, Anim. Nouv. TAmer. Sud, Poiss.: 47 (la riviere des
Amazones); Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw. Cl., 26 (1857):
422 (no locality); Bleeker, 1864, Nat. Verh. Holl. Maatsch. Wetensch., (2) 20: 86 (no
locality)'; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2) 1: 151 (Villa
Bella); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 35 (Amazonas,

1) Certainly Fowler's (1941b, fig. 27) figure of the type of Auchenipterus brachyurus
looks very much like an ordinary A. nuchalis $. It contradicts Eigenmann & Eigenmann’s
(1890: 298) key character that in A. brachyurus the ventrals do not reach the anal, and
the lengths of maxillary as well as mental barbels agree nicely with specimens of
A, nuchalis from Suriname.

i8

ZOOLOGISCHE VERHANDELINGEN 1 32 (1974)

Surinam); Perugia, 1897, Ann. Mus. Genova, 38: 20 (Alto Beni, Bolivia); Goeldi, 1898,
Bol. Mus. Paraense, 2: 461, 479 (Marajo); Eigenmann, 1910, Rep. Princeton Univ. Exp.
Patagonia, 3: 397 (Amazonas, Surinam); A. de Miranda Ribeiro, 1911, Arch. Mus. Nac.
Rio de Janeiro, 16: 397, 478, pi. 54 fig. 1 (Amazonas, Caldeirao, Rio Capin e Paraguay);
Eigenmann, 1912, Mem. Carnegie Mus., 5: 65 (Dutch Guiana, French Guiana); Starks,
I9I3» Stanford Publ. Univ. Ser.: 32 (Para); A. de Miranda Ribeiro, 1914, Comm. Linh.
Telegr. Estr. MattoGrosso ao Amazonas, Annexo 5: 11 (Manaos); Fowler, 1915, Proc.
Acad. Nat. Sci. Philad., 67: 222 (Peruvian Amazon); Fisher, 1917, Ann. Carnegie Mus.,
11: 425 (Para; San Antonio, Rio Madeira; Lagoa de Pamagua); Pearson, 1924, Indiana
Univ. Stud., 11 (64): 15 (Pena Colorado, Huachi, Bolivia); Pearson, 1937, Proc. Calif.
Acad. Sci., (4) 23: no (Beni-Mamore); Fowler, 1939, Proc. Acad. Nat. Sci. Philad.,
91: 231 (Contamana, Ucayali River basin, Peru); Fowler, 1940, Proc. Acad. Nat. Sci.
Philad., 92: 96 (Alto Beni, Pena Colorado, Huachi, Bolivia); Fowler, 1941, Proc. Acad.
Nat. Sci. Philad., 93: 143 (Rio Pamahyba, Therezina, Piauhy); Fowler, 1941, Bol. Mus.
“Javier Prado”, 5: 468 (Peru: Nauta, Contamana); Eigenmann & Allen, 1942, Fish.
W. South America: 121 (Peru: Rio Morona; Iquitos; Gosulimcocha, Rio Morona; Rio
Huallaga, Yurimaguas; Manaos); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool.,
33: 14 (do Paraguai ao Surinam); Menezes, 1949, Rev. Brasil Biol., 9: 489-495, fig. 1-3
(bacia do Rio Parnaiba, Piaui); Bertin & Esteve, 1950, Cat. Types Poiss. Mus. Paris,
5: 16 (Cayenne); Fowler, 1951, Arq. Zool. S. Paulo, 6: 461 (Amazonia, Peru, Bolivia,
Rio Parnaiba, Paraguay, Surinam, Guiana Francesa); P. de Miranda Ribeiro, 1962,
Publ. Avuls. Mus. Nac. Rio de Janeiro, 45: 5 (Manaos, Rio Amazonas; Eiaurupe, Rio
Jurua; Rio Amazonas); Ringuelet, Aramburu & de Ariamburu, 1967, Los Peces Argen¬
tines de Agua Dulce: 274, 319, fig. 18A (distribution in Argentina); P. de Miranda
Ribeiro, 1968, Bol. Mus. Nac. Rio de Janeiro, Zool., 261: 5-7, 9, fig. 9 (no locality).

Euanemus colymbetes ; Muller, 1842, Arch. Anat: 320 (Surinam) ; Muller & Troschel,
1849, Horae Ichth., 3: 11, pi. 1 fig. 2 (Surinam); Bleeker, 1858, Visschen Ind. Arch.,
1: 206 (Surinam).

Euanemus columbetes ; Muller, 1843, Arch. Naturg., 9 (1) : 318 (Surinam).
Euanemus nuchalis ; Bleeker, 1858, Visschen Ind. Arch., 1: 206 (Brasilia) ; Gunther,
1864, Cat Fish. Brit. Mus., 5: 193 (River Capin, Para); Cope, 1878, Proc. Amer. Philos.
Soc., 17: 676, 677 (head streams of the Amazon); Vaillant, 1880, Bull. Soc. Philomat.,
( 7 ) 4* 154 (Calderon (Haute-Amazone)); Kappler, 1881, Hollandisch-Guiana: 167
(Hollandisch-Guiana); Kappler, 1885, Das Ausland, 58: 919 (Hollandisch Guiana);
Kappler, 1887, Surinam: 152 (Surinam).

Euanemus dentatus ; Bleeker, 1858, Visschen Ind. Arch., 1: 206 (Guyana).
Auchenipterus dentatus ; Bleeker, 1864, Nat. Verh. Holl. Maatsch. Wetensch., (2) 20:
86, pi. XVI fig. 1 (Surinama).

Auchenopterus nuchalis ; Berg, 1901, Com. Mus. Nac. Buenos Aires, 1: 301 (Rio de
la Plata, cerca de Buenos Aires; Alto Paraguay (Curumba)).

Auchenipterus demerarae ; Fowler, 1914, Proc. Acad. Nat. Sci. Philad., 66: 266 (Rupu-
nuni River, British Guiana); Caporiacco, 1935, Monit. Zool. Ital., 46: 57 (Great Falls
of Mabura (Demerara)); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 14
(Guiana Inglesa); van der Stigchel, 1946, South American Nematognathi: 103 (no
locality) ; van der Stigchel, 1947, Zool. Meded., 27: 103 (no locality); Boeseman, 1952,
Zool. Meded., 31: 181 (Suriname River near Joden savanna); P. de Miranda Ribeiro,
1968, Bol. Mus. Nac. Rio de Janeiro, Zool., 261: 5, 7, 9, fig. 7 (no locality); Lundberg &
Baskin, 1969, Amer. Mus. Novit, 2398: 6, 26 (no locality).

Auchenipterus ambyiacus ; Fowler, 1941, Bol. Mus. “Javier Prado”, 5: 468 (Peru:
Rio Ampiyaco); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 14 (rio Am-
piyacu); Fowler, 1945, Los Peces del Peru: 66 (Peru: Rio Ampiyaco); Fowler, 1951,
Arq. Zool. S. Paulo, 6: 459 (Alto Amazonas, Peru).

Auchenipterus osteomystax ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool. 33:

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

19

14 (Rio Vermelho, Goiaz); Fowler, 1951, Arq. Zool. S. Paulo, 6: 461 (Goias); P. de
Miranda-Ribeiro, 1962, Publ. Avuls. Mus. Nac. Rio de Janeiro, 45: 5 (Mato-Grosso);
Britski, 1969, Pap. Avuls. Zool., 22: 205 (Rio Vermelho, Santa Rita das Antas, Estado
de Goias).

Auchenipterus brevior\ Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 15
(Guiana Inglesa); van der Stigchel, 1946, South American Nematognathi: 104 (Tuma-
tumari, Lower Potaro, British Guiana); van der Stigchel, 1947, Zool. Meded., 27: 104
(Tumatumari, Lower Potaro, British Guiana); Grey, 1947, Fieldiana, Zool., 32: 124
(British Guiana: Tumatumari, lower Potaro River); Bohlke, 1953, Stanford Ichth.
Bull., 5: 42 (Tumatumari, Lower Potaro River, British Guiana).

Auchenepterus nuchalis; Fowler, 1945, Los Peces del Peru: 66 (Peru: Nauta, Con-
tamana).

Ceratocheilus osteomystax ; P. de Miranda Ribeiro, 1953, Arq. Mus. Nac. Rio de
Janeiro, 42: 391 (no locality).

Material. — One specimen ($), label illegible (RMNH no. 4824), standard length
121 mm, figured as Auchenipterus dentatus by Bleeker (1864), and undoubtedly collected
between the years 1824-1836, near Paramaribo, by H. H. Dieperink. Four specimens
(1 3 , 3 $), January 1946, Suriname Rivier near Jodensavanne (D. C. Geijskes, RMNH
no. 18125), standard length 97, 101, 104 ($), no mm. Eight specimens (2 $, 6 9 ),
3 January 1964, Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25588),
standard length 87 (£), 98, 108 ($), 114, 115, 116, 117, 135 mm. Four specimens (all 9 ),
7 April 1966, Marowijne near Albina (Mees, RMNH no. 25589), standard length 82,
83, 86, 86 mm. One specimen ($ ), 18 April 1967, Loka Loka tabbetje, Marowijne (Nijssen,
ZMA no. 106025), standard length 115 mm. One specimen ($), 24 April 1967, Kamaloea
or Saloea Kreek, French Guiana (Nijssen, ZMA no. 110679), standard length 119 mm.

Characters. — D 1.6 or 1 .7, A 39-45, P I.12, V 13 (i.12), C i.is.i and rudi¬
ments, gill-rakers 32-39 (8-11 + 1 + 23-27). For a general impression of the
appearance of this species, see the generic diagnosis and fig. 3.

There is a dear sexual dimorphism in the species (fig. 3). Males differ
from females in having the maxillary barbels thick and ossified almost to
their tips; these banbds can be moved sideways only, vertical movements
appear to be almost impossible; the first four fully-grown anal rays are modi¬
fied and coalescent, supporting the urogenital tube which nms along their
anterior border to the tip of these rays; the dorsal spine is longer, about 1.5
in predorsal length, as against 1.8 to 2.0 in females; the body appears to be
slightly more slender.

Habitat. — The lower courses of the great rivers (see under the heading
Distribution), where the species appears to be common. It is a free-swimming
species, usually found not far below the surface.

Distribution. — Eastern South America from the Essequibo River south
to the Paraguay-Parana-La Plata basin.

Discussion. — This species was originally described from Brazil. Besides
creating a genus for it, Valenciennes (1840) described a new species, said
to come from Cayenne, as A . dentatus . Very soon afterwards, without refer¬
ence to the works of Spix and Valenciennes, Muller (1842) based Euanemus

20

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

Fig. 3. Anchenipterus nuchalis 9 and S, showing sexual differences (RMNH from
no. 25588). 1.2 X. The arrow indicates the position of the genital pore of the male.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

21

colymbetes on a specimen from Suriname. Bleeker (1864) gave a redescription
and a coloured plate of A. dentatus based on a specimen in Leiden; as locality
he gave Suriname. In the same year Gunther (1864) placed both A . dentatus
and E. colymbetes in the synonymy of A. nuchalis , where they have remained
ever since.

It is perhaps good to mention here that the inclusion of Cayenne and Suri¬
name in the range of A. nuchalis was exclusively based on the type specimens
of A. dentatus and E. colymbetes, the fish figured by Bleeker, and the material
which Kappler (1881) states to have sent to the Stuttgart Museum, no later
material having been recorded from these countries under any of the three
names. Moreover, the locality Cayenne was almost certainly given in error
for Suriname, which means that specimen ZMA no. 110679 constitutes the
first certain record of the genus Auchenipterus for French Guiana.

The first record of the genus from British Guiana was by Eigenmann
(1912) who described A. demerarae, giving as only differential character
a lateral band which, by inference, would not be present in A . nuchalis . There
are not many subsequent records, but A . demerarae appears to have become
accepted as a valid species. Van der Stigchel (1946) brought the Leiden
specimen no. 4824 to A. demerarae , and observed: “The character of the dark
lateral band, given by Eigenmann as the only striking difference between
Auchenipterus nuchalis (Spix) and Auchenipterus demerarae Eigenm., is
very obvious in the here described specimen”. The lateral band is certainly not
obvious now, and I doubt that it was very obvious when van der Stigchel
examined it, as even then it had been in preservative for over a century. There
is, however, some pigment on the flanks, indicative of such a band. Van der
Stigchel, of course, was not aware that the specimen he ascribed to A. deme¬
rarae was the specimen previously described and figured by Bleeker (1864)
under the name A. dentatus , and is probably a topotype of that name.

In the same paper, Eigenmann (1912) described a second new species
from British Guiana, A. brevior , based on the single character of having
mandibular barbels extending to the gill-opening only, as against to the tip
of the pectoral in A . demerarae .

Chronologically the next name was A. ambyiacus Fowler (1915); one speci¬
men of a lot previously identified by Cope (1878) as A. nuchalis , but distin¬
guished by greater depth of body. In the description Fowler gives as body-
depth of A. nuchalis 4^4 to 5, for A. ambyiacus 434? but the illustration shows
a body depth of about 4 1 / 3 in standard length. It is interesting to note that
in the type and only known specimen there was: “From shoulder towards
middle of upper caudal lobe [a] pale dusky-gray streak”, as this is the charac¬
ter on which A. demerarae was based.

22

ZOOLOGISCHE VERHANDELINGEN 132 (l 974 )

A. de Miranda Ribeiro (1918a) differentiated Ceratocheilus osteomystax
as a new genus and species related to Auchenipterus, mainly on the basis of
having: "... barbilhoes maxillares osseos, comprimidos, longos, erecteis ...
Anal falcada, longa, comos raios anteriores muito desenvolvidos, provavel-
mente portadores da papilla genital”. Gosline (1945) synonyxnized the genus
Ceratocheilus with Auchenipterus, but the species has continued to be recog¬
nised.

This historical review leaves us with the present situation, that A. dentatus
and E. colymbetes are considered synonyms of A. nuchalis, and that A. deme-
rarae, A. brevior, A. ambyiacus and A. osteomystax are accepted as valid
species.

It has already been mentioned that A. dentatus agrees with A. demerarae ,
and therefore the last-mentioned name is certainly a synonym of the former.
This means that no more than one species is known from Suriname and
Cayenne.

Van der Stigchel (1946) expressed doubt about the validity of A. brevior.
In my specimens of A . nuchalis from Suriname, there is some individual
variation in the length of the mandibular barbels, from only a little beyond
the base of the pectorals, to nearly the tip of the pectorals; this makes me
suspect that A. brevior also falls within the normal range of variation of A.
nuchalis. Therefore I am of the opinion that A. brevior too, can best be
synonymized with A. nuchalis, although the possibility that it is valid cannot
be entirely dismissed.

A. ambyiacus was exclusively based on body-depth. Subsequently Fowler
(1941a) himself recorded in A . nuchalis a body-depth of from 4Vs to Si and
in Suriname specimens I found a variation of from 4 to 5. As therefore the
only differential character ever claimed for A. ambyiacus is not valid, there
is no ground for maintaining the species.

The characters of Ceratocheilus osteomystax are sexual; it is peculiar that
the existence of sexual dimorphism has not hitherto been recognised in the
species, though Boeseman (1952) made a suggestion in this direction 1 ).

There remains the possibility that Guiana specimens are differentiated
from Amazonian specimens, and as the only available material is from the
Guianas, I have been unable to make a direct comparison. Judging by

1) This section was written before I received P. de Miranda Ribeiro’s (1968a) paper
in which the sexual dimorphism of Auchenipterus is excellently described. Subsequently
I found an even earlier reference: “the males of Auchenipterus have one pair of whiskers,
ossified into great bony horns which swivel forward and upward to remind one of the
horns of an Assyrian bull” (Myers, 1947). This observation, made in a popular article,
does not appear to have penetrated scientific literature.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

23

descriptions, Amazon individuals have slightly higher anal counts. In Suri¬
name specimens the number of anal rays is 39-45, whereas in Amazonia
counts of 45-49 (Fowler, 1940a) and 48 (type of C. osteomystax) have been
recorded. On the other hand, Fisher (1917) gave for three Amazonian speci¬
mens 42, 45 and 47, and Berg (1901) mentioned for the extreme south of
the range (Alto Paraguay and Rio de la Plata), 42-47. In Spix’s (1829)
figure I count 44 rays. It is apparent therefore that, though there appears to
be a slight difference in average number of anal rays, the overlap is too great
to use it as a specific, or even as a subspecific character.

Fowler (1915, in the description of A. ambyiacus, and 1940a) is the only
author ever to have mentioned gill-rakers. His counts (10 + 23 and 14 + 22)
agree with my counts of Suriname specimens.

Genus Psetidauchenipterus Bleeker

Pseudauchenipterus Bleeker, 1862, Atlas Ichth., 2:6 — type by original designation
and monotypy Silurus nodosus Bloch.

Generic diagnosis. — Rather heavily-bodied fishes; head and anterior part
of body roundish in circumference, posterior part compressed; eyes of moder¬
ate size, lateral in position; head with thick skin, covering the remarkably
swollen, “honeycomb-like” frontal tones; nuchal bones hard, covered with
thin skin only, leaving the tone-structure clearly visible; spines strong,
pungent, smooth along their anterior border, serrated behind; barbels well-
developed, longer than in most genera, even inner mandibular barbels reaching
at least to pectoral base (as otherwise only found in Entomocorus ); pectoral
pore far behind pectoral base; ventrals with 8 rays, anal with 2024 rays,
caudal fin well-developed, long, deeply forked. Lateral line complete, with
moderate zig-zags in its anterior part, straight in its posterior part. Sexual
dimorphism exists in the position of the urogenital pore and the shape of the
anal fin, as described under P. nodosus .

Distribution. — Widely distributed in northern and eastern South America.

Discussion. — The diagnosis given above is based on P. nodosus only.
Three other species have been included in the genus : Auchenipterus jequitin -
honhae Steindachner, 1876; Auchenipterus affinis Steindachner, 1876, and
Felichthys flavescens Eigenmann & Eigenmann, 1888. I have not examined
any of these little-known species and doubt that they should really be regarded
as congeneric with P. nodosus. None of them appears to have the “honey¬
comb-like” bone-structure which is such a peculiar feature of P. nodosus. On
the other hand they have similar numbers of dorsal, ventral and anal rays,
and that may be significant.

24

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Gosline (1945: 14) included two further nominal species in the genus.
Pseudauchenipterus guppyi Regan and Pseudauchenipterus paseae Regan,
but the first of these has been generally recognised as a synonym of P.
nodosus (cf. Eigenmann, 1912: 201; Boeseman, i960: 104), and the second
is an evident lapsus for Parauchenipterus paseae Regan, a synonym of Pa -
rauchenipterus galeatus, no Pseudauchenipterus paseae ever having been
described by that author (cf. Regan, 1906).

Pseudauchenipterus nodosus (Bloch)

Silurus nodosus Bloch, 1794, Naturgesch. Ausl. Fische, 8: 35, pi. 368 fig. 1 — Tranque-
bar (errore!) = South America [reference not verified).

Auchenipterus furcatus Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat.
Poiss. (4 0 ed.), 15: 157 — de Cayenne ou de Surinam.

Pseudauchenipterus guppyi Regan, 1906, Proc. Zool. Soc. Lond.: 387, pi. XXIII —
Caroni River, Trinidad.

Pseudauchenipterus nigrolineatus Fowler, 1911, Proc. Acad. Nat. Sci. Philad., 63: 434,
fig. 3 — Pedemales, Venezuela.

Silurus nodosus ; Bloch, 1797, Ichth. Hist Nat. Poiss., 11: 26, pi. 368 fig. 1 (Tran-
quebar) ; Bloch & Schneider, 1801, Syst. Ichth.: 383 (ad Tranquebariam).

Arius nodosus ; Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.), 15: S3 (Tranquebar).

Auchenipterus nodosus; Muller & Troschel, 1849, Horae Ichth., 3: 11 (no locality);
Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw. Cl., 26 (1857) : 424 (Suri¬
nam) ; Bleeker, 1858, Visschen Ind. Arch., 1: 208 (Amer. merid.); Gunther, 1864, Cat.
Fish. Brit. Mus., 5: 194 (British Guiana; Dutch Guiana; French Guiana; Demerara) ;
Goeldi, 1898, Bob Mus. Paraense, 2: 457, 479 (Magoary (Marajo)).

Pseudauchenipterus nodosus ; Bleeker, 1862, Atlas Ichth., 2: 6 (name only); Bleeker,
1863, Ned. Tijdschr. Dierk., 1: 88 (name only); Bleeker, 1864, Nat. Verh. Holl. Maatsch.
Wetensch., (2) 20: 91, pi. XI fig. 1, pi. XIII fig. 6 (Surinama); Eigenmann & Eigen¬
mann, 1891, Proc. U.S. Nat. Mus., 14: 34 (Bahia, Para, Guiana); Eigenmann & Bean,
1907, Proc. U.S. Nat. Mus., 31: 663 (Amazon River between Para and Manaos); Eigen¬
mann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 396 (Bahia; Para; Guiana); A. de
Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 354 (Bahia, Para, Guyanas);
Eigenmann, 1912, Mem. Carnegie Mus., 5: 201, pi. XX fig. 2 (British Guiana: Mahaica;
mud-flats of Demerara River, near Wismar) ; Starks, 1913, Stanford Publ. Univ. Ser.:
32 (Para); Fowler, 1915, Proc. Acad. Nat. Sci. Philad., 67: 222 (Surinam); Fowler,
1915, Proc. Acad. Nat. Sci. Philad., 67: 529 (Caroni River, Trinidad) ; Fisher, 1917,
Ann. Carnegie Mus., 11: 424 (Para; Penedo; Alcobaga; Rio Coite); A. de Miranda
Ribeiro, 1918, Rev. Mus. Paul., 10: 734 (no locality); Fowler, 1940, Proc. Acad. Nat. Sci.
Philad., 91: 231 (Contamana, Peru); Fowler, 1941, Bol. Mus. “Javier Prado”, 5: 468
(Peru (Contamana)); Eigenmann & Allen, 1942, Fish. W. South America: 122 (Conta¬
mana, Peru); Roth, 1943, Fish Life in Brit. Guiana: 118 (British Guiana); Schultz,
1944, Proc. U.S. Nat. Mus., 94: 239 (Venezuela); Gosline, 1945, Bol. Mus, Nac. Rio de
Janeiro, Zool., 33: 13 (Bahia; Para; Guiana; Peru); Fowler, 1945, Los Peces del Peru:
64 (Peru (Contamana)); van der Stigchel, 1946, South American Nematognathi: 101
(Surinam; River Paramaribo; River Demerara; Caribbean Sea); van der Stigchel, 1947,
Zool. Meded., 27: 101 (Surinam; River Paramaribo; River Demerara; Caribbean Sea);
Puyo, 1949, Poiss. Guyane Fr.: 94, fig. 52 (Guyane Frangaise); Fowler, 1951, Arq.
Zool. S. Paulo, 6: 469 (Amazonia, Peru, Para, Rio Sao Francisco, Bahia, Sao Paulo,
Guiana); Price, 1955, J. Agric. Soc. Trinidad & Tobago, Soc. Pap., 863: 14 (Trinidad:

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

25

Caroni river near Piarco) ; Boeseman, 1956, Zool. Meded., 34: 189 (10 miles above outlet
of Suriname River); Boeseman, i960, Studies Fauna Curagao, 10: 104 (Caroni River,
Trinidad); Mago-Leccia, 1967, Soc. Venez. Cienc. Nat Bob, 27: 255 (Venezuela);
Bacon, 1970, The Ecology of Caroni Swamp, Trinidad: 57 (Caroni Swamp, Trinidad).

Euanemus nodosus ; Kappler, 1881, Hollandisch-Guiana: 167 (Hollandisch-Guiana);
Kappler, 1885, Das Ausland, 58: 919 (Hollandisch-Guiana); Kappler, 1887, Surinam: 152
(Surinam).

Felichthys nodosus ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2) 1:
154 (Para; Bahia); Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1: 290
(Para; Bahia).

Pseudauchenipterus guppyi ; Fowler, 1911, Proc. Acad. Nat. Sci. Philad., 63: 433
(Pedernales, Venezuela).

Pseudauchoniptorus nodosus ; Roth, 1943, Fish Life in Brit. Guiana: 69 (British
Guiana).

Material. — Forty specimens from Suriname and Demerara as follows. Two specimens,
11 February 1954, ten miles upstream from mouth of Suriname Rivier (H. W. Lijding,
RMNH no. 21485), standard length 66, 71 mm. Two specimens, 31 March 1966, Suri¬
name Rivier near Paramaribo (W. Vervoort, RMNH no. 26486), standard length 40,
41 mm. One specimen ( 9 ), 10 April 1966, Suriname Rivier near Paramaribo (W. Ver¬
voort, RMNH no. 26487), standard length 225 mm. 17 specimens (10 $, 7 9), 6 Sep¬
tember 1970, Demerara River near Georgetown (J. van der Land, RMNH no. 26488),
standard length 99-157 mm. 11 specimens (6 $,5 9 ), same data (RMNH no. 26489),
standard length 117-183 mm. Nine specimens (3 5 9 , 1 ?), same data (RMNH

no. 26490), standard length 128-193 mm 1 ).

Characters. — D 1 . 6 , A 20-25, P 1-7 (occasionally 1.6 or 1 . 8 ), V 8 (i.7),
C i.15.1, gill-rakers short but fairly numerous, ca. 16 on outer branchial
arch 2 ). The main distinguishing characters have been given in the generic
diagnosis, but a few more will here be added.

Head 3.7-4.5 (usually 3.8-4.0) times in standard length, greatest width of
body (between the cleithra) 4.3-5.2 times in standard length, depth of body
3.6-4-7 times in standard length, predorsal length 2.8-3.35 ^ mes m standard
length. Head nearly cylindrical, about as wide as deep; body compressed,
deeper than wide. Eyes compartively large, about equal in length to snout and
3.8-4.5 (usually about 4.0) times in head; barbels long, the maxillary barbels
reaching at least to the base of the dorsal fin and occasionally as far back
as a vertical through the origin of the anal fin; the outer pair of mental bar¬
bels usually reaches to the tip of the well-developed postcleithral process; the
inner pair of mental barbels is almost or fully as long as the outer pair, but

1) The old material in our collection was listed and described by van der Stigchel
(1946). As adequate fresh material was available, I have only superficially examined
the old specimens.

2) In the forty examined specimens the following aberrant counts were found: DI.5
(one) ; P 1 . 8 , 1.8 (one) ; P 1.7, 1.8 (one) ; P I.7, 1.6 (two); P 1 . 6 , 1.6 (one, the same
specimen that has DT.5); V i.8, i.8 (two), V i.7, i.5 (one); C i.14.1 (one), Ci.i9.i,
divided as follows: i.7.i.n.i (!, one).

26

ZOOLOGISCHE VERHANDELINGEN 1 32 (1974)

does not reach as far backwards as it is implanted in a more anterior position;
even so it reaches at least to the base of the pectoral fin, and often beyond;
the dorsal and pectoral spines are well-developed; the dorsal spine measures
1.0-1.6 times in predorsal length, the pectoral spines are equal in length to
the dorsal spine or up to 10% longer or shorter; the dorsal spine has a smooth
anterior edge; its posterior edge is either smooth or almost smooth or is
provided with small teeth over most of its length; the pectoral spines have a
smooth anterior edge, whereas the posterior edge is provided with a series
of from (11-) 20-36 (-39) well-developed antrorse teeth (there is evidently
an irregular increase in number of teeth with size: the lowest number, n,
was found in a specimen of 40 mm s.l., the highest, 39, in a specimen of
225 mm s.l.); caudal fin deeply-forked with pointed lobes; pectoral pore just
under the middle of the postcleithral process, well behind the implantation
of the pectorals; lateral line well-developed, complete, with irregular arcs or
zig-zags.

Colour. Dark blue-grey above, with yellow or orange flanks and fins;
small and medium-sized specimens have a blackish border to the caudal fin,
which is faint or absent in the largest individuals; on the sides of the body
there are some eight or ten inconspicuous vertical rows of small white spots.

Sexual dimorphism. In males the anterior anal rays are long and the
urogenital pore is at the end of a canal which runs along the anterior edge
of the anal fin and widens into a vesicle towards the tip of the rays. In females
the urogenital pore is situated on the belly, just in advance of the anal origin
and the anterior anal rays are a little shorter, giving the anal fin an almost
straight rather than a falcate outline. The ventral fins tend to be rounded in
females, slightly more pointed in males, but the difference is one of averages
only. I have found no other secondary sexual characters. Fisher (1917:
424-425) appears to have been the first to note and describe the sexual dimor¬
phism of this species.

Habitat. — Evidently an inhabitant of the lower courses and estuaries of
the great rivers, presumably adapted to life in brackish and even salt water.

Distribution. — In Suriname only known from the mouth of the Suriname
Rivier, upstream to Paramaribo, but almost certainly occurring in the estuaries
of all major rivers. Elsewhere recorded from Venezuela, Trinidad, British
and French Guiana, Amazonia, and along the east coast of Brazil south to
Penedo (Fisher, 1917) and Bahia (Eigenmann & Eigenmann, 1890).

Fowler (1951: 460) has recorded this species from as far south as the State
of Sao Paulo, a record apparently based on A. de Miranda Ribeiro, for in the
bibliography of the species Fowler states that A. de Miranda Ribeiro (1918b:
734) had material from Santos (and Sao Joao da Barra). However, Fowler

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

must have misunderstood A. de Miranda Ribeiro’s paper which is a numbered
list of species. No. 158 of this list is Felichthys bagre, of which two specimens
are mentioned, one from Santos and one from S. Joao da Barra. No. 159
is Pseudachenipterus nodosus; a single specimen is listed, without locality,
but with the words: “ja determinado porem sem procedencia”. Fowler has
apparently interpreted this a meaning that this specimen was from the same
locality as the preceding one ( Felichthys bagre), whereas actually it means that
it was without a locality.

Discussion. — The small specimens do not yet have the spongy bone of the
large individuals: the shields of their heads are hard.

Genus Trachycorystes Bleeker

Trachycorystes Bleeker, 1858, Visschen Ind. Arch., 1: 200 — type by original designa¬
tion and monotypy, Auchenipterus trachycorystes Valenciennes 1 ).

Generic diagnosis. — Head broad and depressed, width between the cleithra
almost twice its depth at the same place; bones of head and nape fused to
form one single, heavy and strongly rugose casque, entirely exposed, without
skin; a single small oval fontanel; a wide, horizontal mouth, with the lower jaw
clearly protruding; each jaw with a band of small teeth, none on vomer and
palatines; eyes rather small, lateral; dorsal spine strong, with teeth along its
anterior edge only, the posterior edge smooth; pectoral spines stronger and
much longer than dorsal spine, with teeth along both edges; ventrals with
9 or 10 rays; anal fin of moderate length (ca. 20 rays); posterior part of body
comparatively slender, with a well-developed caudal peduncle and a slightly
forked tail. Lateral line complete, wavy, especially in its anterior part. No
sexual dimorphism has been recorded.

Distribution. — The distribution of this genus is very insufficiently known;
there are records from the Essequi'bo, and Mato Grosso (without exact
locality).

Discussion. — Only two of the many species currently placed in Trachy¬
corystes , appear to belong to the genus; they are T. trachycorystes , the type
of the genus, and T. obscurus. I believe that the one is a synonym of the other
so that the genus, which is extremely little-known, may consist of a single
species. This species is probably widely distributed but rare. The habitat is
presumably the bottom of the large rivers.

Except in a few species with many-rayed ventral fins (genera Auchenipterus
and Pseudepapterus), the number of ventral rays is a good generic character

1) This is the correct reference; in literature, the genus is usually ascribed to Bleeker
(1862) or even Bleeker (1863).

28

ZOOLOGISCHE VERHANDELINGEN 132 (l974)

in the Auchenipteridae. In view of the constancy of this character in the
majority of genera, I was surprised to find that in Trachycorystes a variation
of from six to ten ventral rays has been recorded. The question arose of
whether some of the published counts might not be erroneous. Therefore I
tabulated all species currently recognized, with the number of ventral rays
as given in their original descriptions (Table). Whereas species have been
described with 6, 7, 8, 9 and 10 ventral rays, I established through correspon¬
dence and personal examination that in the genus Trachycorystes as hitherto
defined, the ventral ray numbers actually known to occur, are either 6, or
9 to 10. A priori one would expect this difference to be of generic value, and
examination of material revealed that indeed the 10-rayed species differ in
numerous characters from the 6-rayed species.

List of species currently recognized in the genus Trachycorystes, with their

finray-numbers

name

T. galeatus (Linnaeus, 1766)

T. trachycorystes (Valenciennes, 1840)

T. ceratophysus (Kner, 1858)

T. ohscurus (Gunther, 1863)

T. striatulus (Steindachner, 1876)

T. glaber (Steindachner, 1876)

T, magdalenae Steindachner, 1878

T. insignis Steindachner, 1878

T. isacanthus Cope, 1878

T. brevibarbus Cope, 1878

T. porosus Eigenmann & Eigenmann, 1888

T. analis Eigenmann & Eigenmann, 1888

T. albicrux Berg, 1901

T. amblops (Meek & Hildebrand, 1913)

T. fisheri Eigenmann, 1916
T. leopardinus Borodin, 1927
T. cratensis P. de Miranda Ribeiro, 1937
T. teaguei Devincenzi, 1942
T. coracoideus Eigenmann & Allen, 1942
T. insignis peloichthys Schultz, 1944
T. jokeannae Hoedeman, 1961

number of

number of

recounted

ventral rays

anal rays

number of

as published

as published

ventral rays

6

24

9—10

20

10 x )

6

20

9

19—20

(^~) 10 2)

6

25—27

7-8

23

3 )

6

27—30

6

26

7

22

6 4 )

7

22—23

6 3 )

6

24—25

6

41

6—7

32

6

17—20

—

21—25

6 6)

8

17—18

6 7 )

8

24

8 )

6

23

—

—

9-10 9 )

6

26—28

6

20

*) Personal recount from holotype. 2 ) Personal recount from syntype. 3 ) Type (or
types?) lost (Kahsbauer, in Iitt.). 4 ) Bohlke, in litt. 5 ) Fowler (1915 : 222). 6 ) Eigen¬
mann (1922, pi. V). 7 ) Baskin, in litt. 8 ) Type not traced; position not clear: by its

describer stated to be near T. trachycorystes. 9 ) Personal recount from syntypes;
judged a synonym of Aucheniptichthys thoracatus .

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

29

Type of the genus Trachycorystes is T. trachycorystes, a species with 9-10
ventral rays. For the species with 6-rayed ventrals, the generic name Parau-
chenipterus is available, of which P . galeatus is the type. It is unfortunate but
inevitable that the familiar binomen Trachycorystes galeatus, in use since (as
it now appears mistakenly) the Eigenmanns (1888) synonymized Parau -
chenipterus with Trachycorystes, has to be abandoned, and that the generic
name Trachycorystes becomes restricted to a single little-known species.

Possibly, Trachycorystes cratensis A. de Miranda Ribeiro (1937) is a
second member of this genus, as it was described as “muito proximo de 7 \
trachycoristes According to P. de Miranda Ribeiro (1953: 407), the type
was in the Museu Nacional, reg. no. 947, but curiously, in the catalogue (P.
de Miranda Ribeiro, 1962b) it is not mentioned. For reasons explained in the
introduction, I have made no further attempts to trace this specimen.

Trachycorystes trachycorystes (Valenciennes) (fig. 4, pi. 1)

Auchenipterus trachycorystes Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat.
Poiss. (4 0 ed.), 15: 159, pi. 437 — no locality (“nous le croyons de Bresil”).

Trachycorystes typus Bleeker, 1862, Atlas Ichth., 2:6 — nomen novum for Auchenip¬
terus trachycorystes Valenciennes, given to avoid tautonomy.

Auchenipterus obscurus Gunther, 1863, Ann. Mag. Nat. Hist., (3) 12: 442 — the
Essequibo.

Trachycorystes trachycorystes ; Bleeker, 1858, Visschen Ind. Arch., 1: 208 (Amer.
merid.); Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2) 1: 154 (no
locality) ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1: 276 (habitat:?);
Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34 (no locality); Eigenmann,
1910, Rep. Princeton Univ. Exp. Patagonia, 3: 396 (Brazil) ; Gosline, 1945, Bol. Mus.
Nac. Rio de Janeiro, Zool., 33: 12 (Brasil); Fowler, 1951, Arq. Zool. S. Paulo, 6: 476
(Brasil?).

Auchenipterus trachycorystes ; Gunther, 1864, Cat. Fish. Brit. Mus., 5 : J 95 (Brazil).
Auchenipterus obscurus ; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 195 (Essequibo).
Trachycorystes obscurus; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) 1: 154 (no locality) ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci.,
1: 275 (Essequibo); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34
(Essequibo); Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 396 (Esse¬
quibo) ; Eigenmann, 1912, Mem. Carnegie Mus., 5: 200, pi. XVII fig. 2 (no locality =
Essequibo); Beebe, 1925, Zoologica, 6: 120 (Kartabo); Gosline, 1945* Bol. Mus. Nac.
Rio de Janeiro, Zool., 33: 12 (Essequibo); Alexander, 1964, Proc. Zool. Soc. Lond., 142:
420 (British Guiana); Alexander, 1966, Proc. Zool. Soc. Lond., 148: 90 (British
Guiana).

Trachycorystes trachycoristes ; A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de
Janeiro, 16: 366 Para ou Matto Grosso).

Trachycoristes trachycoristes ; P. de Miranda Ribeiro, 1962, Publ. Avuls. Mus. Nac.
Rio de Janeiro, 45: 11 (Estado de Mato-Grosso).

Material. — One stuffed specimen without data (MNP no. A 9422), standard length
ca. 310 mm, holotype of the species. One specimen, no date, Essequibo River (Ehrhardt,
BM no. 1864-1-21: 13-14), standard length 164 mm, syntype of Auchenipterus obscurus.

3 °

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

31

One specimen, undated, Estado de Mato-Grosso, without exact locality (L. Travassos,
MRJ no. 1501), standard length ca. 70 mm.

Characters. — D 1 . 6 , A 20, P 1.6 or I.7, V 9-10, C i.i5.i and rudiments.
The main characters of this species have been given in the generic diagnosis
and additional characters are listed in the discussion. A general impression of
its apperance may be obtained from fig. 4 and pi. 1.

Distribution. — The only exact localities are the Essequibo River with its
tributaries the Cuyuni and the Maraballi, but the species is probably widely
distributed in Amazonia. The type of the species is a mounted individual,
taken by the French Armies from the Royal Cabinet in Lisbon, and there¬
fore presumed by Valenciennes to have come from Brazil. According to A.
de Miranda Ribeiro (1911: 367) : “...a regiao de procedencia sejam os estados
do Para ou Matto Grosso, ambos visitados por Alexandre Rodrigues Fer¬
reira, o fornecedor do Museu de Lisboa, de animaes do Brasil”. Specimen
MRJ no. 1501 confirms this provenance.

Discussion. — In the original description of Auchenipterus obscurus, no
reference was made to T . trachycorystes (cf. Gunther, 1863), but differential
characters between the former and the latter were given by Gunther (1864)
as follows:

T. obscurus
D I.5, P I.7 or 1.8

bones of head and neck rather
coarsely granulated

the maxillary barbels extend to, or
somewhat beyond, the extremity of
the humeral process

the humeral process extends
backwards beyond the middle of the
pectoral spine

the dorsal spine is considerably
shorter than the pectoral spine

caudal fin slightly emarginate

T. trachycorystes
D 1 . 6 , P 1.6

bones of head and neck finely
granulated

the maxillary barbels are not longer
than the head

the humeral process extends
backwards to, or a little beyond, the
extremity of the pectoral spine

the dorsal spine is nearly as long as
the pectoral spine

caudal fin truncate

In their key to the species of Trachycorystes , Eigenmann & Eigenmann
(1890: 273-274) separated T. obscurus and T. trachycorystes widely on the
basis of the shape of the tail, ascribing to the former a “caudal emarginate”,
to the latter a “caudal obliquely truncate or rounded; lower jaw longer”. It

32

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

should be noted that the Eigenmanns had not examined material, their
descriptions were copied from those given by Valenciennes and Gunther, and
that Gunther had examined T. obscurus only and not the unique type of T.
trachycorystes in Paris. A direct comparison between the types of the two
nominal species had therefore not previously been made. The results of such
a comparison, made by me, will be summarized here.

The dorsal fin of T. obscurus has six rays, not five. The sixth ray is small
and weak; it is only half as long as the fifth ray but appears to be clearly
separated from it and is not just a branch of the fifth. The posterior part of
the dorsal fin of T. trachycorystes is broken off, but it looks as if there have
been six rays, as already described by Valenciennes. In the pectoral fins I
count I.754 and 1.6 respecively, but the last rays are fine, and in T. trachy¬
corystes the fins are partly broken, and further covered with a thick layer
of paint and varnish, which may well conceal one ray. There is certainly no
argument here for retention of two different species.

Both T. obscurus and T. trachycorystes have the casque without skin and
coarsely granular. Even though some of the coarseness has been smoothed
over by varnish in T. trachycorystes , there is hardly any difference between
the two.

The maxillary barbels of T. trachycorystes reach to the base of the
pectorals, but their tips are broken. The barbels of T. obscurus are very
slender distally and if broken at the level of the pectoral base, would Offer
about the same aspect as those of T. trachycorystes. Anyway, proof that any
difference in this character exists cannot be given.

In both specimens the tip of the humeral process extends to below the horn
of the nuchal plate; in T. obscurus the pectoral spine reaches to well beyond
the tip of the humeral process, in T. trachycorystes only a little beyond the tip
of the humeral process.

Even though its tip is broken off and missing, it can be seen that the dorsal
spine of T. trachycorystes is considerably shorter and weaker than the
pectoral spine; adding the estimated length of the missing part it would have
been about three-quarters the length of the pectoral spine, exactly the same
proportion as found in T. obscurus.

T. trachycorystes has the tips of the caudal fin broken off, and it is clear
that the median rays, which are almost entire, have been shorter than the
outer rays; it is likely therefore that T. trachycorystes had the same shape of
caudal fin as T. obscurus. The small specimen MRJ no. 1501 has the caudal
fin obliquely truncate rather than emarginate.

Both specimens have a broad and somewhat protruding lower jaw, with
identical dentition.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

33

Apart from the characters already discussed, the figure in Valenciennes
(1840: pi. 437), which represents the specimen very well, suggests that T.
trachycorystes has a much lower anal fin than T. obscurus, but here again,
the explanation is that in the former the rays are broken.

Finally, in T. trachycorystes the body is larger and heavier, compared with
the head, fins, and other solid or semi-solid structures.

The general agreement between T. trachycorystes and T. obscurus is such
that I have little hesitation in uniting them. Such differences in proportions as
exist can easily be explained, partly by the fact that the specimen of T. trachy¬
corystes is twice the size (standard length) of the specimen of T. obscurus,
partly by the fact that the specimen of T. trachycorystes has evidently been
stuffed very well: the body is almost certainly much heavier, and perhaps also
longer, than it would have been if it had been preserved in alcohol, and other
proportions also appear to have been distorted. This leaves Trachycorystes
with a single species.

The species appears to be rare: except for the specimen in the Museu
Nacional, Rio de Janeiro, listed by P. de Miranda Ribeiro (1962b), and also
examined by me, the species had been recorded from its type only, collected
probably before the end of the eighteenth century. The nominal species
Auchenipterus obscurus was based on two syntypes, collected before 1863.
Not much additional material had been mentioned in literature, for the speci¬
men recorded by Steindachner (1915) under the name of Trachycorystes
obscurus belongs to a different species, Auchenipterichthys longimanus.
Actually, however, the British Museum collection contains some further speci¬
mens: two from the R. Cuyuni, British Guiana (Carter, BM no. 1934.9.12:
375-376), standard length ca. 250-300 mm, and one from the Maraballi River,
a tributary of the lower Essequibo (R. M. Alexander, BM no. 1961.1.23.1),
standard length ca. 100 mm (information received from Mr. G. Howes).
From this it looks as if in the Essequibo basin the species is not excessively
rare. It is fortunate that Alexander has deposited one of his specimens in
the British Museum, as otherwise the identity of the material he used for his
anatomical studies might have been queried. I have failed to trace Beebe’s
(1925) ichthyological material from Kartabo (it is not in AMNH), but see
no reason to doubt his record.

Genus Auchenipterichthys Bleeker

Auchenipterichthys Bleeker, 1862, Atlas Ichth., 2:7 — type by original designation and
monotypy, Auchenipterus thoracatus Kner.

Generic diagnosis. -— Moderately heavy-bodied; head not depressed, almost
as high as wide; bones of head and nape exposed or covered with very thin

3

34

ZOOLOGISCHE VERHANDELINGEN 1 3 2 ( 1974 )

skin; fontanel large, elongate; eyes large, lateral in position; mouth not
particularly wide; jaws equal, each jaw with a comparatively narrow band of
teeth, none on vomer and palatines; pectoral pore on, or a little behind, a
vertical through the posterior end of the pectoral base, just below the well-
developed postcleithral process; ventrals with 9 or 10 rays; anal fin of
moderate length (20-25 rays), caudal peduncle not slender, but not so short
and compact as in Parauchenipterus. Lateral line complete, wavy, giving off
short branches. Sexual dimorphism in shape of the anal fin and the caudal
fin has been described from one species.

An exposed coracoid process, forming a bony plate below each pectoral
spine, has sometimes been regarded as an important generic character, but it
is found in only one of the two species (which otherwise are extremely
similar), and therefore can have no place in the generic diagnosis.

This genus is close to Parauchenipterus, but differs by its slightly more
slender body, larger fontanel, larger eyes, and by having nine or ten ventral
rays.

Distribution. — Confined to the Amazon basin, where widely distributed.

Discussion. — The genus contains two very similar species.

Auchenipterichthys thoracatus (Kner)

Auch[enipterns] thoracatus Kner, 1858, Sitzb. Akad. Wiss. Wien, M'athem.-Naturw.
Cl., 26 (1857) : 425, pi. VII fig. 22 — Rio Guapore.

Auchenipterus thoracicus Gunther, 1864, Cat. Fish. Brit. Mus., 5: 194 — emendation
of A. thoracatus .

Trachycorystes coracoideus Eigenmann & Allen, 1942, Fish. W. South America: 120 —
Iquitos.

Auchenipterichthys thoracatus ; Bleeker, 1862, Atlas Ichth., 2: 7 (name only); Bleeker,
1863, Ned. Tijdschr. Dierk., 1: 89 (name only) ; Eigenmann & Eigenmann, 1888, Proc.
Calif. Acad. Sci., (2) 1: 154 (Coary; Hyavary); Eigenmann & Eigenmann, 1890, Occ.
Pap. Calif. Acad. Sci., 1: 282 (Coary; Hyavary); Eigenmann & Eigenmann, 1891, Proc.
U.S. Nat. Mus., 14: 34 (Solimoens and tributaries); Eigenmann, 1910, Rep. Princeton
Univ. Exp. Patagonia, 3: 396 (Solimoens and tributaries (Coary; Hyarary; Guapore));
A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 373, fig. 140 (Guapore,
Javary, Coary); (pt.) Fisher, 1917, Ann. Carnegie Mus., 11: 424 (Maciel, Rio Guapore;
San Joaquin; Bastos, Rio Alegre, into R. Guapore); Pearson, 1937, Proc. Calif. Acad.
Sci. (4) 23: no (Mamore basin); Fowler, 1940, Proc. Acad. Nat. Sci. Philad., 92: 96
(Bolivia: San Joaquin); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 13
(Solimoes e afluentes); Fowler, 1951, Arq. Zool. S. Paulo, 6: 458, fig. 486 (Amazonia,
Bolivia); P. de Miranda Ribeiro, 1962, Pap. Avuls. Mus. Nac. Rio de Janeiro, 45: 4
(Uypiranga, Rio Negro, Estado do Amazonas; Rio Quixito, embocadura do Javary,
Estado do Amazonas); P. de Miranda Ribeiro, 1968, Bol. Mus. Nac. Rio de Janeiro,
263: 8, pi. VII (no locality).

Trachycorystes coracoideus ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 13
(Iquitos); Fowler, 1945, Los Peces del Peru: 64 (Peru (Iquitos)) ; Fowler, 1951, Arq.
Zool. S. Paulo, 6: 473 (Alto Amazonas, Peru).

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

35

Material. — Two specimens, 1922, Iquitos (Morris, CAS ex IUM no. 15974), standard
length 103, 106 mm, syntypes of Trachycorystes coracoideus. One specimen, 23 July
1943, Yahnas Yacu, near Pebas, Peru (W. G. Scherer, SU no. 58656), standard length
67 mm.

Characters. — D 1 . 6 , A 23-25, P 1 . 8 , V 9 (i.8), C i.is.i and rudiments.
Head 3.6-3.85 times in standard length, greatest width of body (between the
cleithra) 4.04.3 times in standard length, height of head at the same place
0.8 times its width, depth of body 3.8-4.2 times in standard length, predorsal
length 2.75-2.95 times in standard length, dorsal spine 1.4-1.45 in predorsal
length, pectoral spine 1.1-1.3 in predorsal length.

This species is extremely similar to the next one, but has a slightly more
slender (more compressed) head, a greater predorsal length, the horns of the
n;uchal plate are much broader, the postcleithral process is thicker with a
blunter tip, the pectoral spines, when pressed against the body reach to or a
little beyond the ventral origin, the number of anal rays is higher, and finally
there is the conspicuous character of the exposed bone-plates below the
pectorals.

Kner (1858) described a sexual difference in shape of the caudal fin,
which is obliquely truncate in females, and with an elongated upper lobe in
males. He also mentioned that the majority of his specimens consisted of
females. Eigenmann & Eigenmann (1890: 284) described the anal fin as
rounded or truncate in females, concavo-convex in males. Judging by these
characters, the three specimens examined by me are females.

Colours. Preserved specimens are dark brown above, pale below. On the
upper part of the body, beginning on the opercles, and continued to below
the adipose fin, there are ten to twelve vertical rows of pale dots, two to four
to a row. These dots have no connection with the lateral line system; they
are also higher up on the body, differently arranged and larger than the spots
of A. longimanus, and provide an additional character to distinguish between
the two species.

Distribution. — Upper course of the Amazon and its southern tributaries.
The most eastern locality from where this species has been recorded is
Coary = Rio Coari (Eigenmann & Eigenmann, 1890).

Discussion. — As a matter of routine, to complete the table on page 28, I
Ichecked up on the number of ventral rays of Trachycorystes coracoideus
Eigenmann & Allen, and to my surprise was informed by Dr. Eschmeyer that
they are 9-10. In the description of T. coracoideus, Eigenmann & Allen
failed to give a single finray-count, but as they stressed its close relationship
with T. galeatus, I had assumed that T. coracoideus belonged to the group with

36

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

6 -rayed ventrals. Examination of two syntypes subsequently revealed the
synonymy here indicated.

Auchenipterichthys longimanus (Gunther)

Auchenipterus longimanus Gunther, 1864, Cat. Fish. Brit. Mus., 5: 195 — River Capin.

Auchenipterichthys longimanus ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad.
Sci., (2) 1: 154 (Manes, Rio Madeira; Cameta); Eigenmann & Eigenmann, 1890, Occ.
Pap. Calif. Acad. Sci., 1: 284 (Maues, Rio Madeiro; Cameta); Eigenmann & Eigen¬
mann, 1891, Proc. U.S. Nat. Mus., 14: 34 (Southern tributaries of the Amazon); Eigen¬
mann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 396 (Southern tributaries of the
Amazon); A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 372 (Rio
Capim, Maues, Madeira, Cameta); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool.,
33: 13 (afluentes do Amazonas); Fowler, 1951, Arq. Zool. S. Paulo, 6: 458 (Para, Rio
Madeira).

Trachycorystes obscurus ; Steindachner, 1915, Denkschr. Akad. Wiss. Wien, Mathem.-
Naturw. Kl., 93: 80 (Mundung des Rio Negro); (pt.) Fowler, 1951, Arq. Zool. S. Paulo,
6: 474 (Amazonas).

Auchenipterichthys thoracatus \ (pt.) Fisher, 1917, Ann. Carnegie Mus., n: 424
(Manaos).

Material. — One specimen, no date, River Capin, Para (pres. Stevens, BM no. 1849-
11-8: ?), standard length 135 mm, syntype of the species. One specimen, 1913, Miindung
des Rio Negro (J. D. Haseman, MV no. 43371), standard length 129 mm.

Characters. — D 1.6 or I.7, A 20-22, P 1.6 to 1 . 8 , V 9-10 (i.8-i.9), C i.15.1
and rudiments. Differential characters from A. thoracatus have been given
under that species. There is no trace of exposed bone-plates below the pec¬
torals.

Description. — A normally-shaped member of the family; snout depressed,
posterior part of head almost as high as deep; body compressed, especially
the caudal peduncle which is much deeper than wide.

Head 3.75-4.2 times in standard length, greatest width of body (between
the cleithra) 4.24.5 times in standard length, height of head at the same place
0.8-0.9 times width, depth of body 3-8-3-9 times in standard length, predorsal
length 3.2-3.35 times in standard length.

Snout depressed, much wider than high, but height of head increasing
backwards, to almost as high as wide at the level of the cleithra; profile from
snout to base of dorsal fin ascending, almost straight; bones of head and nape
covered with very thin but strongly pigmented skin, through which the bone
structure is clearly visible; depression containing the fontanel elongate, more
than three times as long as wide, reaching to very near the tip of the snout;
horns of nuchal plate thin, strongly bent outwards and downwards; width of
mouth about half width of body; each jaw with a band of about five irregular
rows of small teeth; no teeth on vomer or palatines; eyes large, their greatest
diameter no more than 3.5-3.8 times in head, and 24-2.7 times in bony inter-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

37

orbital; nostrils placed in a rectangle, the distance separating the two of each
pair about 2.4 times the distance separating the anterior from the posterior
ones; anterior nostrils placed on the vertical part of the snout, posterior
nostrils placed horizontally, between the eyes; maxillary barbels thin, reaching
to the end of the well-developed postcleithral process, which is to half-way
along the pectoral spines, or a little shorter, reaching to half-way along the
ptocess; outer pair of mental barbels about half their length, not quite to
(base of pectoral spine; inner pair, which is placed in a far more anterior
position, only about one-fourth of the length of the outer pair, and no more
than half an eye’s diameter in length.

Dorsal fin with one spine and six or seven rays; the spine strong, slightly
curved backwards, its length 1.4-1.5 times in predorsal length, finely serrated
along its entire anterior border, the small teeth becoming sparse towards its
tip and appearing in a double row towards its base, and with about 25 slightly
larger teeth, diminishing in size towards the base, along its posterior border;
spine with a soft tip or filament; spine with filament 1.2-1.3 times in pre¬
dorsal length; the first ray scarcely shorter than spine with filament, the
following ones successively smaller, and the last ray weak and no more than
one-third of the length of the first ray.

Anal fin with 20-22 rays, its base about as long as the head and 1.2-1.5 times
in predorsal length; the anterior rays equal to postorbital part of head, the
last ones much shorter, no more than one-third of the anterior ones, and less
than an eye’s diameter; the anterior rays concealed in thick skin; urogenital
pore at about two-thirds from the base on the anterior border of the fin.

Pectoral fins with one strong spine and six to eight rather weak rays; the
slightly curved spine is serrated along both sides over its whole length, along
the anterior edge there are ca. 35 fairly small teeth, directed outwards, along
the posterior edge there are ca. 30 somewhat larger teeth, directed inwards;
the spine measures about 1.15-1.2 times in the predorsal length, one-fourth
longer than the dorsal spine, and equal in length to dorsal spine with filament;
when pressed against the body it falls just short of the ventral origin. The
first ray is almost as long as the spine, but the succeeding rays diminish
rapidly.

The ventrals have one simple and eight or nine divided rays; they are
implanted entirely behind the dorsal fin, rounded in outline with the third,
fourth and fifth rays longest; they measure about 2.2 times in predorsal
length, and fall a little short of the fleshy base of the anal fin.

The caudal fin is shallowly emarginate, with fifteen divided rays, flanked
on each side by one fully developed simple ray, and some shorter simple rays
and rudiments.

38

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

The adipose fin is well-developed, rounded, opposite the posterior section
of the anal fin and reaching about as far backwards as the anal fin.

The lateral line consists of small irregular arcs giving off short branches;
it is complete, continued on to the base of the caudal fin.

Colour. Body and fins of MV no. 43371, which appears to have retained
much of its natural colour, entirely blackish grey without markings, except
the ventral surface from chin to anus which is white, unpigmented. In the
older specimen, the black has been replaced by dull-brown. Both specimens
show the pores of the lateral line system, which are unpigmented, as an
irregular series of small but by no means inconspicuous white dots.

Distribution. — Mainly middle- and lower course of the Amazon and its
southern tributaries from the Rio Purus (Cameta) to the Rio Capim (rather
than Capin), near Para. There is one record from Nazareth, Peru (see dis¬
cussion). Although A. thoracatus and A. longimanus have clearly different
ranges, there appears to be a large area of co-occurrence in the upper-middle
Amazon.

Discussion. — The Vienna specimen (MV no. 43371) agrees very well
with the syntype of A. longimanus, a species which was evidently unknown
to Steindachner (1915), who reported his specimen under the name Trachy -
corystes obscurus, but nevertheless he observed already some of its characters:
“Das uns vorliegende Exemplar aus dem Amazonasgebiete unterscheidet sich
nur unbedeutend von den typischen Exemplaren des britischen Museums, und
zwar durch die schlankere, langliche (nicht runde) Form der Stirnfontanelle,
die viel geringere Lange der vorderen Mentalbarteln und das Vorkommen
von 10 Strahlen in der Ventrale sowie von 7 Gliederstrahlen in der x. Dor-
sale”.

Fisher (19x7) listed under the name A. thoracatus a fish from Manaos. As
all other records of that species are from farther upstream, and as A . longi¬
manus was already known to occur near Manaos, I considered that the record
required verification. From Mr. Schoknecht I received the following infor¬
mation: “I checked the specimen of Auchenipterichthys ... This was CM
6782 and is now FM 57799. As you suspected, the specimen was not identified
correctly. From the information that you sent, it appears that it is A. longi¬
manus. The only other A. longimanus that we have in our collection is one
lot of two small (40-50 mm s.l.) specimens from Nazareth, Peru (Osgood,
10 Sept. 1912)”.

Genus Parauchenipterus Bleeker

Parauchenipterus Bleeker, 1862, Atlas Ichth., 2:7 — type by original designation,
Silurus galeatus Bloch — Sihirus galeatus Linnaeus.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

39

Generic diagnosis. — A genus of generally short, heavy, round-bodied
fishes; the head not much depressed, only a little wider than deep; bones of
head and nape covered with skin which is thin and often transparant in small
specimens, but thick and heavily pigmented in large individuals; pectoral pore
above middle of pectoral base; fontanel small, smaller than eyes, oval, behind
the middle of the eyes; a moderate to long anal fin (17 to 40 rays), 6 ventral
rays, caudal peduncle short, caudal fin obliquely truncate, obliquely rounded,
or slightly emarginate; eyes small; lower jaw slightly protruding. Sexual
dimorphism in this genus, first mentioned by Eigenmann & Eigenmann
(1890: 272) and fully described by R. von Ihering (1937), is usually slight,
but P. insignis and P. fisheri show a pronounced sexual dimorphism in shape
of head, maxillary barbels, dorsal and anal fins (Miles, 1947: 74; Eigenmann,
1922: pi. V). Except for the first one, these are the same sexual characters
as found in Auchenipterus nuchalis ; they support the opinion that the genera
Auchenipterus and Parauchenipterus are related, and should certainly not be
placed in different families.

Distribution. — Northern South America from Panama eastwards, and
the whole of eastern South America east of the Andes, south to the Uruguay-
Parana-La Plata basin.

Discussion. — On a previous page I have given a list of nominal species
hitherto included in Trachycorystes; with the exception of T. trachycorystes ,
r. obscurus, T. coracoideus and possibly T. cratensis, all these species are
referable to Parauchenipterus . This would make Parauchenipterus a fair-sized
genus (16 species and one subspecies), but it is unlikely that all described
species are valid: on the following pages evidence will be given that T. glaber
and T, jokeannae are synonyms of P. galeatus, and Fisher (19x7) considered
7 \ striatulus to be “undoubtedly identical” with the same species. A revision
based on adequate material would probably lead to a reduction of the number
of species to four or five, with in addition perhaps a few valid subspecies.

From Suriname only a single species is known, which is common and
Widely distributed in the country, and has a wide range in eastern South
America.

Parauchenipterus galeatus (Linnaeus)

Silurns galeatus Linnaeus, 1766, Syst. Nat., ed. 12, 1: 503 — in America australi, here
restricted to Suriname (see Discussion).

Auchenipterus maculosus Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat.
Poiss. (4 0 ed.), 15: 161 — Cayenne (but probably Suriname).

Aucheniptems immaculatus Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat.
Poiss. (4 0 ed.), 15: 162 — Cayenne (but probably Suriname).

Auchenipterus punctatus Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat.
Poiss. (4 0 ed.), 15: 163 — probablement du Bresil.

40

ZOOLOGISCHE VERHANDEL1NGEN 13 2 ( 1974 )

Auchenipterus robustus Gunther, 1864, Cat. Fish. Brit. Mus., 5: 197 — Demerara.

Auchenipterus glaber Steindachner, 1876, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw.
Cl., 74 (1) : 655 — Demerara.

Parauchenipterus paseae Regan, 1906, Proc. Zool. Soc. Lond.: 387, pi. 24 (not 23) —
Caroni River, Trinidad.

Trachycorystes jokeannae Hoedeman, 1961, Bull. Aquatic Biol., 2: 138 — French
Guiana: lie de Cayenne, ruisseaux du Rorota.

Siluriis galeatus ; Gmelin, 1789, Syst. Nat., ed. 13, 1 (3): 1397 (in America australi);
Bloch, 1797, Ichth. Hist. Nat. Poiss., 11: 30, pi. 369 fig. 1 (rAmerique meridionale);
Bloch & Schneider, 1801, Syst. Ichth.: 384 (in America australi).

Pimelodus galeatus; La Cepede, 1803, Hist. Nat. Poiss., 5: 97, 114 (rAmerique meri¬
dionale).

Auchenipterus maculosus; Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw.
Cl., 26 (1857): 425 (Marabitanos) ; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 196 (Suri¬
nam, Essequibo); Vaillant, 1880, Bull. Soc. Philomat., (7) 4: 154 (Calderon (Haute-
Amazone)); Kappler, 1885, Das Ausland, 58: 918 (Hollandisch Guiana); Kappler, 1887,
Surinam: 151 (Surinam); Pellegrin, 1899, Bull. Mus. Hist. Nat. Paris, 5: 158 (I’Apure
et ses affluents, Venezuela).

Parauchenipterus galeatus ; Bleeker, 1862, Atlas Ichth., 2: 7 (name only); Bleeker,
1863, Ned. Tijdschr. Dierk., 1: 88 (name only) ; Bleeker, 1864, Nat. Verh. Holl. Maat-
sch. Wetensch., (2) 20: 45 (Surinama).

Auchenipterus galeatus ; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 196 (British Guiana,
Guiana); Boulenger, 1896, Trans. Zool. Soc. Lond., 14: 28 (Matto Grosso).

Anchenipterus galeatus; Peters, 1877, Mber. Akad. Wiss. Berlin: 470 (Calabozo,
Venezuela).

Euanemus maculosus ; Kappler, 1881, Hollandisch-Guiana: 167 (Hollandisch-Guiana).

Trachycorystes glaber ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2)
1: 154 (no locality) ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1: 275
(Demarara); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34 (Demarara) ;
Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 395 (Demarara) ; Eigen¬
mann, 1912, Mem. Carnegie Mus., 5: 198 (Demerara); Gosline, 1945, Bol. Mus. Nac.
Rio de Janeiro, Zool., 33: 11 (Demerara); Hoedeman, 1961, Bull. Aquatic Biol., 2: 138
(no locality).

Trachycorystes galeatus ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) 1: 155 (Pernambuco; San Gongallo; Rio San Francisco, below the Falls; Tabatinga;
Rio Puty); Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1: 279 (Per¬
nambuco; San Gongallo, Rio San Francisco, below the falls; Tabatinga; Teffe; Rio
Puty); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 15: 34 (Rio das Velhas
to the Orinoco); Eigenmann & Bean, 1907, Proc. U.S. Nat. Mus., 31: 663 (Amazon
River between Para and Manaos); Eigenmann, 1910, Rep. Princeton Univ. Exp. Pata¬
gonia, 3: 396 (Rio Velhas to the Orinoco and Paraguay); Eigenmann, 1912, Mem. Car¬
negie Mus., 5: 198 (various localities in British Guiana) ; Starks, 1913, Stanford Publ.
Univ. Ser.: 31 (Para); A. de Miranda Ribeiro, 1914, Comm. Linh. Telegr. Estr. Matto-
Grosso ao Amazonas, 15, Annexo 5: 10 (Manaos) ; Fowler, 1914, Proc. Acad. Nat. Sci.
Philad., 66: 266 (Rupununi); Bertoni, 1914, Fauna Paraguaya: 8 (Paraguay); Fowler,
1915, Proc. Acad. Nat. Sci. Philad., 67: 222 (Surinam; Rupununi River); Fowler, 1915,
Proc. Acad. Nat. Sci. Philad., 67: 329 (Caroni River, Trinidad); Popta in Benjamins &
Snelleman, 1914-1917, Enc. Ned. West-Indie: 685 (Suriname; Rio Velhas tot de Orinoco
en Paraguay); Fisher, 1917, Ann. Carnegie Mus., n : 423 (Para; Manaos; Santarem;
Maciel, Rio Guapore; Rio San Francisco, Bolivia; San Joaquin, Bolivia; Rio Jauru,
into R. Paraguay; Puerto Suarez, Rio Paraguay; Arequa, Paraguay; Uruguayana;
Penedo; Barreiras, Rio Sao Francisco; Joazeiro; Rio das Velhas; Munez Freire, Rio
Itapemerin; Campos; Lagoa Feia; Sao Joao da Barra; Rio Itapicuru, Queimadas; Rio

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

41

Itapicuru, Timbo); A. de Miranda Ribeiro, 1918, Rev. Mus. Paul., 10: 734 (Rio
S. Francisco; Itaqui, Rio Grande do Sul); Pearson, 1924, Indiana Univ. Stud., 11 (64):
15, 54 (Lake Rogoagua, Rio Beni basin, Bolivia); Pearson, 1937, Proc. Calif. Acad. Sci.,
(4) 23: no (Beni and Mamore basins; Paraguay basin); Fowler, 1940, Proc. Acad. Nat.
Sci. Philad., 92: 96 (Bolivia: San Joaquin, Puerto Suarez, L. Rogoagua); Fowler, 1941,
Proc. Acad. Nat. Sci. Philad., 93: 143 (Lago Papary, Rio Grande do Norte; Guara-
miranga, Ceara; Pianco, Parahyba; Forteleza, Ceara); Schultz, 1944. Proc. U.S. Nat.
Mus., 94: 239 (Venezuela: no original material); Gosline, 1945, Bol. Mus. Nac. Rio de
Janeiro, Zool., 33: 12 (do Rio das Velhas ao Orinoco e Paraguai); van der Stigchel,
1946, South American Nematognathi: 99 (Surinam; Brazil; Lama Stop-off, British
Guiana) ; van der Stigchel, 1947, Zool. Meded., 27: 99 (Surinam; Brazil; Lama Stop-off,
British Guiana) ; Puyo, 1949, Poiss. Guyane Fr.: 98 (Cayenne?); Bertin & Esteve, 1950,
Cat. Types Poiss. Mus. Paris, 5: 17 (Cayenne (Guyane)); Fowler, 1951, Arq. Zool.
S. Paulo, 6: 473 (Amazonia, Peru, Bolivia, Ceara, Rio Sao Francisco, Rio Parana,
Uruguay, Guiana, Venezuela, Trinidad); Boeseman, 1952, Zool. Meded., 31: 181 (Nanni
creek, Doksen savanna; Coropina creek, Republiek) ; Boeseman, 1953, Zool. Meded., 32: 9
(Vierkinderen, Republiek); Lijding, 1959, Surinaamse Landb., 7: 73 (Suriname); Boe¬
seman, i960, Studies Fauna Curasao, 10: 103 (Trinidad); Lowe, 1964, J. Linn. Soc.
Lond., Zool., 45: no, 112, 119, 121, 141, 143 (Rupununi District, British Guiana);
Ringuelet, Aramburu & de Aramburu, 1967, Los Peces Argentinos de Agua Dulce:
278 (distr. in Argentina) ; McConnell, 1967, Timehri, 43: 67 (Rupununi) ; Ovchynnyk,
1968, Zool. Anz., 181: 253 (Ecuador: Rio Pana Yacu, upper Rio Napo, Prov. Napo-
Pastaza); Marlier, 1968, Cad. Amazonia, 11: 53, 55 (lac du Rio Preto da Eva); Lund-
berg & Baskin, 1969, Amer. Mus. Novit., 2398: 6, 27 (no locality).

Trachycorystes robustus ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2)
1: 136 (no locality); Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1:
274, 281 (Demarara) ; Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34
(Demarara) ; Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 396 (Deme-
rara).

Trachycoristes galeatus; A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de
Janeiro, 16: 370 (Rio S. Francisco, Amazonas e tributarios, Marajo, Orenoco e ilha da
Trinidade, Matto Grosso); Magalhaes, 1931, Monogr. Braz. Peix. Fluviat.: 102, fig. 39
(Para, Rio S. Francisco, Amazonas e Matto-Grosso).

Material. — 195 specimens from Suriname as follows. Three specimens, 27 February
1964, rapids Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25590), standard
length 67, 113, 139 mm. One specimen, about 20 April 1964, Suriname Rivier near
Aloesoebanji (P. Leentvaar, RMNH no. 25658), standard length 36 mm. One specimen,
early May 1964, lower course Sara Kreek (P. Leentvaar, RMNH no. 25653), standard
length 115 mm. Four specimens, 23 June 1964, Suriname Rivier near Aloesoebanji
(Boeseman, RMNH no. 25655), standard length 25, 37, 38, 39 mm. Two specimens,
9 July 1964, Soekroewatra, Brokopondo Meer (P. Leentvaar, RMNH no. 25591), standard
length 52, 67 mm. 24 specimens, 10 July 1964, Brokopondo Meer, ca. 2 km E. of Kabel
along railway to Abontjima (Boeseman, RMNH no. 25592), standard length 32-46 mm.
84 specimens, same data (Boeseman, RMNH no. 26530), standard length 36-110 mm.
Three specimens, 21 July 1964, tributary of Gran Kreek ca. 12 km from its mouth
(Boeseman, RMNH no. 25593), standard length 32, 32, 40 mm. Three specimens,
20 March 1965, Suriname Rivier near Brokopondo (Mees, RMNH no. 25659), standard
length 155, 156, 160 mm. Two specimens, 24 March 1965, Brokopondo Meer, along rail¬
way from Kabel to Abontjima (Mees, RMNH no. 25594), standard length 114, 136 mm.
Two specimens, 2 April 1965, Gansee, Brokopondo Meer (Mees, RMNH no. 25595),
standard length 139, 147 mm. One specimen, 7 April 1965, Para Kreek (Mees, RMNH
no. 25596), standard length 39 mm. Five specimens, 8 April 1965, Gansee, Brokopondo

42

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Meer (Mees, RMNH no. 25597), standard length 142, 145, 145, 147, 154 mm. Four
specimens, same data (Mees, RMNH no. 25598), standard length ill, 1x7, 125, 143 mm.
One specimen, 20 April 1965, Para Kreek (Mees, RMNH no. 25599), standard length
39 mm. One specimen, 24 April 1965, Para Kreek (Mees, RMNH no. 25600), standard
length 173 mm. One specimen, 27 April 1965, Para Kreek (Mees, RMNH no. 25651),
standard length 44 mm. Three specimens, 22 June 1965, Gansee, Brokopondo Meer (Mees,
RMNH no. 25764), standard length 128, 133, 146 mm. Twelve specimens, 24 June 1965,
Brokopondo Meer, along railway between Kabel and Abontjima (Mees, RMNH no.
25643), standard length 109-157 mm. Five specimens, 1 November 1965, Brokopondo
Meer, along railway between Kabel and Brownsweg (Mees, RMNH no. 25644),
standard length 118, 119, 122, 129, 134 mm. Four specimens, 19 December 1965, Com-
pagnie Kreek (Mees, RMNH no. 25645), standard length 59, 67, 81, 96 mm. Four speci¬
mens, same data (Mees, RMNH no. 25656), standard length 87, 106, 107, 129 mm. Nine
specimens, same data (Mees, RMNH no. 25660), standard length 61-104 mm. One
specimen, 3 March 1966, creek near Zanderij (Mees, RMNH no. 25646), standard length
144 mm. One specimen, 6 March 1966, Brokopondo Meer, along railway between Kabel
and Brownsweg (Mees, RMNH no. 25647), standard length 166 mm. Six specimens,
12 March 1966, tributary of Mama Kreek near Berg-en-Dal (Mees, RMNH no. 25648),
standard length 94, 94, 96, 102, 102, 104 mm. Five specimens, 15 March 1966, Compagnie
Kreek (Mees, RMNH no. 25649), standard length 79, 95, 104, 109, 139 mm. Two speci¬
mens, same data (Mees, RMNH no. 25657), standard length 27, 52 mm. One specimen,
8 June 1966, Gran Kreek, 41 km S. of Affobakka (W. Vervoort, RMNH no. 26531),
standard length 72 mm.

Extra-limital material. One specimen, received in 1911. Ipu, Ceara (MV no. 43372,
ex Mus. Goeldi), standard length 129 mm. One specimen, 3 November 1957, ruisseaux
du Rorota, lie de Cayenne, French Guiana (J. Gery, ZMA no. 102371), standard length
90 mm, holotype of Trachycorystes jokeannae.

Characters. — D 1 . 6 , A 2025, P 1.6 to 1 . 8 , V 6 (i.5), C 14-15 (divided
rays only). The main characters have been given in the generic diagnosis;
in this species the caudal fin is obliquely truncate, with the upper portion the
longer. No sexual dimorphism.

Colours of small specimens mottled brown or blackish on a pale brown or
cream coloured background, or with more or less clearly defined longitudinally
elongated blackish dots. Large specimens very dark, blackish brown, irre¬
gularly mottled with black. Caudal fin usually with a black cross-bar.

Distribution. — Northern South America from the Orinoco basin east¬
wards, the whole Amazon basin, and eastern South America south to the
Parana-La Plata basin.

Discussion. — Linnaeus (1766: 503) based this species entirely on the
description and figure given by Seba (1758: 85, pi. 29, fig. 7). Unfortunately,
Seba did not provide his specimen, which was well-described and reasonably
well-figured, with a locality. Parauchenipterus galeatus is widely distributed
in South America, but as about the middle of the eighteenth century, because
of monopolies, the Spanish and Portuguese possessions in South America
would have been difficult of access for Dutch traders, it is likely that Seba,
who lived in Amsterdam, received his specimen from one of the Dutch pos-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

43

sessions: Essequibo, Demerara, Berbice, and Suriname (see also Valencien¬
nes, 1840: 380). As Parauchenipterus galeatus may show geographical varia¬
tion in its huge range, I regard it as desirable to restrict its type-locality. A
restriction to Suriname ensures not only that my material is correctly named,
but appears also fully justified on the basis of the historical considerations
given above. I regard specimens from the lower course of the Suriname
Rivier and its tributaries as topotypical.

Auchenipterus maculosus Valenciennes, generally accepted as a synonym
of Parauchenipterus galeatus, lastly by Bertin & Esteve (1950: 17) who
examined the type^specimen, was based on a specimen received from Leiden;
as type-locality Valenciennes gave Cayenne, but at the time there was little
zoological material from Cayenne in Leiden. There were, and are, on the
other hand several specimens of Parauchenipterus galeatus from Suriname in
our collection, received from H. H. Dieperink who lived in Suriname from
ca. 1816-1836. I note that Bertin & Esteve (1950) give as locality of the type
of Auchenipterus maculosus : “Cayenne (Guyane)”. Therefore it is likely
that the specimen was originally labelled as from Guyane, and that Valen¬
ciennes has changed this to Cayenne, not realising that with Guyane actually
Suriname was meant. In this connection I want to draw attention to Valen¬
ciennes’ statement (in Cuvier & Valenciennes, 1840: 128) that the Leiden
museum has received specimens of Pimelodus sebae (= Rhamdia quelen)
from Cayenne, for actually all our old specimens are from Suriname. Van
der Stigchel (1946: 50) lists nine specimens received from Dieperink, they,
or at least some of them, are syntypes of Rhamdia sebae . For more extensive
notes on this subject, see Boeseman (1972).

As briefly mentioned on a previous page, Dr. Kahsbauer has been unable
to trace the type (or types ?: Steindachner did not state on how many speci¬
mens his description was based) of Auchenipterus glaber in the Vienna
museum. The reasons why I have, nevertheless, ventured to place the name
in the synonymy of P. galeatus are the following: i° except for the high
ventral ray count, Steindachner’s description fits large specimens of P.
galeatus ; 2 0 ventral rays have often been miscounted in descriptions (cf.
table on p. 28); 3 0 a specimen of P. galeatus from Ceara, received on loan
from Vienna (MV no. 43372) bore the identification Trachycorystes glaber,
an identification presumably made by Steindachner himself; 4 0 Demerara
is ichthyologically reasonably well known, and P. galeatus is the only member
of its genus that has ever been found there, or anywhere else in the Guianas.

The type of Trachycorystes jokeannae is a normal specimen of P. galeatus,
a species not mentioned in its description. The number of anal rays is 21
(i.20), not 20 (i.19) as Hoedeman wrote, and comes within the normal range

44

ZOOLOGISCHE VERHANDELINGEN I 32 ( 1974 )

of variation of P. galeatus. In view of the presumed synonymy of A. glaber
with P. galeatus it is significant that Hoedeman did regard his T. jokeannae
as closely related to A. glaber, and also that he failed to comment on the
discrepancy in ventral ray numbers.

Genus Centromochlus Kner

Centromochlus Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw. Cl., 26 (1857) :
430 — based on Centromochlus megalops Kner and Centromochlus aidopygius Kner;
type by subsequent selection, Centromochlus megalops Kner (Bleeker, 1862: 7).

Generic diagnosis. — D I.5 or l-S l A, A 7 or 8, P I.7 to I.12, V 6 (i.5), C
i.15.1, plus rudiments. In general appearance not unlike Tatia; mouth inferior
and narrow, its lower jaw longer than broad; teeth smaller than in Tatia, in
each jaw in a narrow band, bending backwards to follow the shape of the jaw,
the premaxillary band in the middle very narrow or interrupted (fig. 5);
fontanels situated in a longitudinal depression reaching from the tip of the
snout backwards; eyes large to very large; barbels of variable length, the
maxillary barbels either longer or shorter than in Tatia ; anterior part of the
body relatively long, distance from tip of snout to base of dorsal spine
2.7-3.25 times in standard length; horns of nuchal plate bent upwards and
inwards; pectoral pore over middle of pectoral base; dorsal and pectoral
spines strong and very long, pungent, either without teeth and hooks, and
only slightly rough, or the pectoral spines serrated on the inside only, pectoral
spines reaching to ventrals or nearly so, 2.6-3.2 times in standard length;
lateral line consisting of small arcs; anal fin normal, without a sexual differ¬
ence in shape; ventrals implanted from just behind the middle of the body
to as far backwards as about two-thirds of the body length from the tip of
the snout; caudal fin forked, its lobes with acute tips; caudal peduncle very
distinct. No sexual dimorphism, except in position of the urogenital pore,
as indicated in the discussion of C. existimatus.

Fig. 5. Shape of bands of teeth in Centromochlus heckelii (left, specimen of 68 mm s.l.)
and Tatia intermedia (right, specimen of 68 mm s.l.). 6 X.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

45

All specimens examined, of both species, are in alcohol light brown above,
pale below, usually without any trace of markings or a pattern, but sometimes
with oblique dark bands; fins hyaline.

The most conspicuous characters by which Centromochlus may be separated
from Tatia, are the long pungent spines, and the distinct caudal peduncle.

Distribution. — The genus Centromochlus is widely distributed in Ama¬
zonia and occurs also in the Orinoco basin, but is not yet known from the
Guianas.

Discussion. — The genus Centromochlus as here defined contains only
two, possibly three, species. All other species hitherto placed in Centromochlus
are referable to Tatia. For further particulars, see the discussion of the genus
Tatia .

Fig. 6. Right hand pectoral spines seen from above. Centromochlus heckelii (top, speci¬
men of 68 mm s.l.), and Tatia intermedia (bottom, specimen of 68 mm s.L). 4 X.

Centromochlus heckelii (Filippi) (figs. 5-8, pi. 2)

Auchenipterus Heckelii Filippi, 1853, Rev. Mag. Zool., (2) 5: 167 — Rio Napo.
Cent[romochlus] megalops Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw. Cl.,
26 (1857) : 430 — Bogota.

Centromochlus Steindachneri Gill, 1871, Proc. Acad. Nat. Sci. Philad., 1870: 95 —
no locality given, but perhaps from the Maranon or Napo Rivers, Peru (see discussion).

Centromochlus megalops; Kner, 1858 (1859?), Arch. Naturg., 24 (1): 350 (no locality) ;
Bleeker, 1863, Ned. Tijdschr. Dierk., 1: 89 (name only).

Centromochlus heckelii; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 197 (Bogota; Rio
Napo); Cope, 1878, Proc. Amer. Philos. Soc., 17: 677 (Peruvian Amazon); (pt.) Eigen-
mann & Eigenmann, 1888, Proc Calif. Acad. Sci., (2) x: 156 (Villa Bella; Obidos; Para;
Tabatinga; Lago Alexo; Hyavary); (pt.) Eigenmann & Eigenmann, 1890, Occ. Pap.
Calif. Acad. Sci., 1: 267 (Villa Bella; Obidos; Para; Tabatinga; Lago Alexo; Hyavary);
(pt.) Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34 (Amazonas and
tributaries); Boulenger, 1898, Trans. Zool. Soc. Lond., 14: 422 (Rio Jurua, Brazil);
Regan, 1905, Proc. Zool. Soc. Lond.: 190 (Rio Negro and its tributaries); Eigenmann,
1910, Rep. Princeton Univ. Exp. Patagonia, 3: 395 (Amazons and tributaries); (pt.)
A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 359, 472, pi. 52 fig. 2

46

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

(Villa-Bella, Obidos, Para, Tabatinga, Lago Aleixo, Javary, Caldeirao, (Amazonas),
Juruia); Starks, 1913, Stanford Publ. Univ. Ser.: 31 (Para); (pt.) Fowler, 1915, Proc.
Acad. Nat. Sci. Philad., 67: 221 (Manaos harbor, Brazil) ; Arnold & Ahl, 1936, Fremdl.
Siisswasserfische: 248, fig. (der Amazonenstrom und seine Nebenfliisse sowie die nord-
lichen Zufliisse des Parana); Pearson, 1937, Proc. Calif. Acad. Sci., (4) 23: no
(Mamore basin); Fowler, 1940, Proc. Acad. Nat. Sci. Philad., 92: 96 (Bolivia: Rio
Mamore, San Joaquin) ; (pt.) Fowler, 1941, Bol. Mus. “Javier Prado”, 5: 466 (Pebas);
Eigenmann & Allen, 1942, Fish. W. South America: 117 (Amazon system: Rio Morona;
Rio Huallaga, Yurimaguas; mouth Rio Pacaya; Iquitos; Rio Ucayali, near Orellana);
Schultz, 1944, Proc. U.S. Nat. Mus., 94: 240 (Venezuela); Gosline, 1945, Bol. Mus. Nac.
Rio de Janeiro, Zool., 33: 10 (Amazonas e tributarios; rio Apure, Venezuela); (pt.)
Fowler, 1945, Los Peces del Peru: 62 (Pebas); Myers, 1947, Aquarium J., 18: 14 (the
Amazon); (pt.) Fowler, 1951, Arq. Zool. S. Paulo, 6: 463 (Amazonia, Bolivia, Peru,
Venezuela); (pt.) P. de Miranda Ribeiro, 1962, Publ. Avuls. Mus. Nac. Rio de Janeiro,
45: 9 (Manaos. Rio Amazonas, Estado do Amazonas; Mercado de “Ver-o-Peso”, Belem,
Estado do Para; Belem, Estado do Para; Rio Jurua, Municipio de Eiaurupe, Estado do
Amazonas; Borba, Estado do Amazonas); Mago-Leccia, 1967, Soc. Venez. Cienc. Nat.
Bol., 27: 255 (Venezuela); Marlier, 1968, Cad. Amazonia, 11: 56 (Rio Preto).

Centromochlus Heckelii ; Vaillant, 1880, Bull. Soc. Philomat, (7) 4: 154 (Calderon
(Haute-Amazone)) ; Pellegrin, 1899, Bull. Mus. Hist Nat. Paris, 5: 158 ( 1 *Apure et
ses affluents); Tortonese, 1940, Bol. Mus. Torino, 48: 136 (Rio Napo).

Centromochlus steindachneri ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) 1: 156 (no locality) ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci.,
1: 268 (Upper Amazon) ; Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34
(Maranon); Fowler, 1941, Bol. Mus. “Javier Prado”, 5: 466 (Pebas); Eigenmann &
Allen, 1942, Fish. W. South America: 118 (Pebas); Gosline, 1945, Bol. Mus. Nac. Rio
de Janeiro, Zool., 33: 10 (Maranon) ; Fowler, 1945, Los Peces del Peru: 62 (Peru (Rio
Maranon)); Fowler, 1951, Arq. Zool. S. Paulo, 6: 464 (Alto Amazonas, Rio Maranon).

Centromochlus heckeli ; Fisher, 1917, Ann. Carnegie Mus., 11: 422 (Santarem; San
Joaquin, Bolivia; Para; Rio Mamore; Manaos).

Centromochlus hekelii ; A. de Miranda Ribeiro, 1914, Comm. Linh. Telegr. Estr.
Matto-Grosso ao Amazonas, Annexo 5: 10 (Manaos).

Centromochlas steindachneri ; Lundberg & Baskin, 1969, Amer. Mus. Novit., 2398:
6, 26 (no locality).

Material. — One specimen, 1847, Bogota (MV no. 47360), standard length 63^ mm,
syntype of C. megalops . One specimen, no date, Villa Bella (received from Steindachner
in 1874, presumably Thayer Exp., MV no. 47342), standard length 55 mm. One speci¬
men, 1865/1866, Jatuarana (Navez, Thayer Exp., MCZ no. 7635), standard length 65 mm.
One specimen, Para (MCZ no. 7636), standard length 44 mm 1 ). One specimen, Hyavary
(MCZ no. 7623), standard length 69 mm. One specimen, Tabatinga (MCZ no. 7632)*
standard length 68 mm. One specimen, Obidos (MCZ no. 7627), standard length 55 mm.
Three specimens, Lago Alexo near Manaos (S. V. R. Thayer, Thayer Exp., MCZ
no. 7634), standard length 62, 64, 67 mm. Two specimens, 1865/1866, Teffe (L. Agassiz,
Thayer Exp., MCZ no. 36188), standard length 60, 63 mm. Eight specimens, Villa Bella
(MCZ no. 7730), standard length 5734-62 mm. Ten specimens, Villa Bella (MCZ
no. 7630), standard length 51-66 mm. One specimen, Villa Bella (presumably Thayer Exp.,
ZMK no. 16, ex MCZ), standard length 62 mm. One specimen, 1893/1895, TApure, Vene¬
zuela (F. Geay, MNP no. 98-33), standard length 70 mm. One specimen, 1908, Para
(Goldi, MV no. 47350), standard length 56 mm. One specimen, 1908, Manaos (Comissao
Rondon, MRJ no. 979), standard length 70^ mm. Two specimens, 19 September 1909,

1) There is little doubt that this and following specimens which bear no date or col¬
lector's name, originate from the Thayer Expedition, 1865/1866.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

47

Rio Mamore, Brazil-Bolivia (J. D. Haseman, FM no. 57804), standard length 60, 76 mm.
Two specimens, no date — December 1909, Santarem (J. D. Haseman, CAS no. 24799),
standard length 54, 56 mm. Five specimens, April 1913, harbour of Manaos (E. A. Smith,
ANSP no. 73154), standard length 63, 65, 65, 68, 68 mm. Two specimens, August 1920,
Rio Ucayali near Orellana, Peru (W. R. Allen, CAS no. 24804), standard length 54J/2,
86 mm. Two specimens, August 1920, mouth of Rio Pacaya, Peru (W. R. Allen, CAS
no. 24802), standard length 57, 65 mm. Two specimens, September 1920, Iquitos, Peru
(W. R. Allen, CAS no. 24803), standard length 71, 76mm. Two specimens, October
1920, Rio Morona, Peru (W. R. Allen, CAS no. 24800), standard length 61, 90 mm.
One specimen, November 1920, Rio Huallaga, Yurimaguas, Peru (W. R. Allen, CAS
no. 24801), standard length 75 mm. Two specimens, 22 April 1924, Tocantins R. near
Arumatheua (C. Ternetz, CAS no. 6687), standard length 85, 85 mm. One specimen,
2 June 1924, Tapajos (C. Ternetz, CAS no. 6686), standard length 72 mm. Five specimens,
June 1924, Santarem, Rio Tapajos (C. Ternetz, CAS no. 6676), standard length 59, 60,
61, 63, 64 mm. One specimen, no date = ca. June 1924, Amazon below Santarem (C.
Ternetz, CAS no. 6683), standard length 68 mm. Five specimens, 1943, Borba (A. Parko,
MRJ no. 9413), standard length 68, 72, 73, 73, 75 mm. One specimen, 1944, Borba (A.
Parko, MRJ no. 9414), standard length 68 mm. Seven specimens, 1951, Municipio de
Eiaurupe, Rio Juruia (J. C. de Melo Carvalho, MRJ no. 5884), standard length 48-60 mm.
Three specimens, 1958, Mercado do “Ver-o-Peso”, Belem (L. Travassos and F. Pires,
MRJ no. 9029), standard length 64, 65, 70 mm. One specimen, 24 October 1963, Parana
da Eva, Amazonas (G. Marlier, KBIN no. 16429), standard length 60 mm. Two speci¬
mens, no date, South America (Beal and Steare, UMMZ no. 56134: “C. steindachneri ”),
standard length 66, 66 mm. Two specimens, no date, Pebas, Peru (no collector, CAS
no. 24805: “C. Steindachneri), standard length 66 l />, 70 mm. One specimen, no date,
Belem (C. Estevao de Oliveira, MRJ no. 955), standard length 86 mm.

Characters. — D 1 .5, A 7 or 8, P I.7 to I.9, V 6, C i.15.1 and rudiments
(once i.i4.i); predorsal length 2.7-3.0 times in standard length; distance from
tip of snout to base of ventrals 1.65-1.8 times in standard length; maxillary
barf>els long, reaching to well beyond origin of dorsal fin; the shallow depres¬
sion containing the fontanels does not reach to the supraoccipital; horns of
nuchal plate laterally rounded; posterior edge of pectoral spine rough, but
only in the distal part of the spine do the protuberances assume the shape of
antrorse teeth; adpressed pectoral spines reaching ventrals, the length of
these spines about equal to predorsal length; adipose fin implanted before
a vertical through the posterior border of the adpressed anal fin.

Colour. Preserved specimens are plain light brownish on the dorsal,
yellowish white on the ventral surface. According to Arnold & Ahl (1936)
the species is in life dark grey to greenish grey above, the flanks olive green,
belly yellowish white, evidently without any markings.

Distribution (fig. 7). — The species is widely distributed in the Amazon
basin, from the mouth of the Amazon (Para) to Peru (many localities) and
Bolivia (San Joaquin). Also known from Venezuela (Orinoco drainage) and
Colombia. The only record from Colombia concerns the two syntypes of C.
megalops which are supposed to be from Bogota. But Bogota is a town high
in the Andes, in the immediate surroundings of which Centromochlus is

48

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

Fig. 7. The distribution of Centromochlus heckelii, from material examined (except for
the type-locality, Rio Napo, which is from literature). 1, Apure. 2, Bogota. 3, Rio Morona.
4 Yurimagas. 5, Orellana. 6, Rio Pacaya. 7, Rio Napo. 8, Iquitos. 9, Pebas. 10, Eiaurupe.
11, Tabatinga (and mouth of Hyavary). 12, Teffe. 13, Manaos. 14, Borba. 15, Rio Preta
da Eva. 16, Villa Bella. 17, Obidos. 18, Santarem. 19, Para (Belem). 20, Mamore.

21, Arumatheua.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

49

unlikely to occur, and about the middle of last century it was also an impor¬
tant trade-centre for natural history specimens. The nearest rivers, the Mag¬
dalena and Meta, flow north and north-east respectively, and ultimately into
the Caribbean, but it is very well possible that the specimens were obtained
in one of the rivers of the upper Amazon drainage. Although the ichthyo¬
fauna of the Magdalena and Meta rivers has been studied (Eigenmann,
1922; Miles, 1947), C. heckelii has not been recorded from them. On the
other hand, the species is already known from the Rio Apure, a tributary of
the Orinoco (as is the Meta), and therefore it is likely that the species also
occurs in the Meta (cf. Eigenmann, 1922: 220).

Discussion. — Of the two names here listed as synonyms, C. megalops has
been generally recognized as such, but C. steindachneri requires some discus¬
sion.

First about the type-locality of C. steindachneri . In the original description,
Gill (1871) did not give a locality for his specimen, but because of the caption
of the article in which it was described: “fishes ... from the Maranon, or
Upper Amazon, and Napo Rivers”, later authors have made the Maranon the
type-locality of the species. It is very doubtful that this is correct, for in the
introduction to his paper, Gill writes: “In an expedition to the Andes of
Ecquador and Peru, and thence across the continent of South America, under
the command of Prof. James Orton, a considerable zoological collection was
formed, and the fishes being submitted to the writer for determination, the
following appeared to be undescribed”. The fact that with the majority of
the described species no locality is given, and in one instance a specimen
was: “taken in the river Napo or Maranon”, suggests strongly that the speci¬
mens were not labelled, and as according to the sentence just quoted, the
Orton expedition crossed the continent of South America, the material of
Centromochlus steindachneri may have been taken anywhere in the Amazon
basin (see also Fowler, 1940a: 219).

As in the original description no characters are given which would distin¬
guish this species from C . heckelii (indeed. Gill states that it is most nearly
related to C. megalops, a synonym of C. heckelii ), I have tried to trace the
type, but have been unsuccessful. Neither in the Academy of Natural Scien¬
ces, nor in the United States National Museum could it be found (Bohlke, in
litt.; Gibbs, in litt.).

Since its description, only Eigenmann & Allen (1942) appear to have
recorded material under the name C. steindachneri, all other mention in
literature being just references to the original description. I have examined
Eigenmann & Allen's specimens (now CAS) and two more examples,
originally apparently belonging to the same lot, from UMMZ. These speci-

4

50

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

mens tend to have the groove containing the fontanel rather short, and the
ventrals are placed far backwards (distance from tip of snout to base of
ventrals 1.6-1.65 times in standard length as against 1.65-1.8 in the other
material), but in every other respect they agree with C. heckelii, and it is my
opinion that they fall within the normal range of variation of C. heckelii.

Centromochlus existimatus species nova (figs. 8, 9)

Centromochlus heckelii ; (pt.) Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) I : 156 (Villa Bella; Obidos; Gumpa; Manacapuru); (pt.) Eigenmann & Eigenmann,
1890, Occ. Pap. Calif. Acad. Sci., 1: 267 (Villa Bella, Obidos; Gurupa; Manacapuru);
(pt.) Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34 (Amazonas and
tributaries); (pt.) Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 395 (Ama¬
zonas and tributaries); (pt.) A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de
Janeiro, 16: 359 (Villa-Bella, Obidos, Gumpa, Manacapuru); (pt.) Fowler, 1915, Proc.
Acad. Nat. Sci. Philad., 67: 221 (Manaos harbor); Fowler, 1940, Proc. Acad. Nat. Sci.
Philad., 91: 231 (Contamana, Ucayali River); (pt.) Fowler, 1941, Bol. Mus. “Javier
Prado'*, 5: 466, 467, fig. 23 (Contamana); (pt.) Fowler, 1945, Los Peces del Peru: 62,
63, fig. 23 (Contamana); (pt.) P. de Miranda Ribeiro, 1962, Publ. Avuls. Mus. Nac.
Rio de Janeiro, 42: 9 (Borba, Estado do Amazonas).

Material. — Six specimens, 1865/1866, Gurupa (L. Agassiz, Thayer Exp., MCZ
no. 7631), standard length 44-55 mm. Two specimens, 1865/1866, Obidos (Thayer Exp.,
MCZ no. 7628), standard length 53, 57 mm. One specimen, 1865/1866, Manacapuru
(Thayer Exp., MCZ no (7629), standard length 54 mm. Five specimens, 1865/1866, Villa
Bella (Thayer Exp., MCZ no. 7630), standard length 52, 53, 55, 57, 57 mm. Seven
specimens, 1865/1866, Villa Bella (Thayer Exp., MCZ no. 7730), standard length 55-
60 mm. One specimen, presumably 1865/1866, Villa Bella (presumably Thayer Exp., BM
1889-11-14: 39-40, ex MCZ), standard length 57 mm. Two specimens, 1865/1866,
Jatuarana (Navez, Thayer Exp., MCZ no. 7635), standard length 63, 63 mm. One
specimen, presumably 1865/1866, Villa Bella (presumably Thayer Exp., UZMK, ex
MCZ), standard length 55 mm. One specimen, no date = December 1909, Santarem
(J. D. Haseman, CAS no. 24799), standard length 57 mm. One specimen ($), April
19x3, harbour of Manaos (E. A. Smith, ANSP no. 73154), standard length 66^2 mm,
holotype of the species. One specimen, 3 June 1924, Turute, Amazonas (C. Ternetz, CAS
no. 6689), standard length 39 mm. One specimen, June 1924, Tapajos near Santarem
(C. Ternetz, CAS no. 6676), standard length 64 mm. One specimen, July-August 1937,
Contamana, Ucayali R. basin, Peru (W. C. Morrow, ANSP no. 108489), standard length
60 mm. Two specimens, 1943, Borba, Estado do Amazonas (A. Parko, MRJ no. 9413),
standard length 64, 66 mm. One specimen, no data (MRJ no. 9415), standard length
56 mm.

Characters. — D 1.5^2, A 7 or 8 (mostly iii.5), P I.io to I.i2, V 6 (i-5),
C i.i5.i and rudiments (once i.i6.i), predorsal length 2.9-3. x times in stan¬
dard length, distance from tip of snout to base of ventrals 1.8-1.95 times in
standard length, hence ventrals slightly more anterior in position than in
C. heckelii ; maxillary barbels short, reaching from a little beyond base of P
to just to a vertical through origin of D; the shallow depression containing the
fontanels reaches the supraoccipital; horns of dorsal plate laterally with a

s.l. 63 mm), predorsal part of body viewed from above. Note the difference in extent
of the depression containing the fontanels, and the difference in length of the maxillary

barbels. 3 X.

52

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

sharp angle; posterior edge of the pectoral spines with about 30 well-
developed antrorse serrae; adpressed pectoral spines just reaching ventrals,
the spines shorter than the distance from tip of snout to origin of D (about
four-fifths of this distance); adipose fin implanted on or behind a vertical
through the posterior border of the adpressed anal fin.

The shallow central groove containing the fontanels is continued to the
supraoccipital. In both species, C. heckelii and C. existimatus, there are two
small round fontanels, separated by a bridge, and behind the posterior fon¬
tanel the frontals are broadly joined, but whereas in C. heckelii this area is
not depressed and on a level with the bones forming the outer aspect of the
cranium, in C. existimatus it continues as a shallow depression to the supra¬
occipital. When the skin is in place, in C . heckelii it still looks as if the fren¬
tals are broadly connected in front of the supraoccipital, but in C. existimatus
it looks as if they are not connected, as the joint is sunk in and covered by
skin. I do not understand Eigenmann & Eigenmann’s (1890: 267) descrip¬
tion : “Fontanel elongate, widest behind”, as in all specimens examined by me
(and these include the material studied by Eigenmann & Eigenmann), there
are two fontanels as described above. In some large specimens, however, the
anterior fontanel is somewhat elongated and the posterior one reduced in
size. The Eigenmanns may have examined such a specimen and have over¬
looked the small posterior fontanel, or they may have described the shape
of the skin-covered depression, and not that of the actual fontanels.

The difference in number of pectoral rays between this species and the
preceding one is obvious, but the difference in dorsal rays is also diagnostic:
C. heckelii has five evenly spaced rays, whereas C. existimatus has six rays,
of which the fifth and the sixth originate very close together on the same
pterygiophore; in the material studied this difference was without exception.
The anal fin has two or three simple and five divided rays; as the first simple
ray is small and very slender it may have been present in some specimens in
which I failed to find it, making A 8 the normal formula.

Colour. For a description of colour and pattern I refer to the generic
description.

Distribution (fig. 9). — On present evidence this species is widely
distributed along the main stream of the Amazon, from Gurupa to Conta-
mana. Doubtless its actual distribution embraces most of the Amazon basin.

Discussion. — The specimen described and figured by Fowler (1940a,
1941b, 1945b) under the name C. heckelii actually is C. existimatus (speci¬
men ANSP no. 108489, examined). Fowler's figure is a reasonable presen¬
tation, showing well the short maxillary barbel and many-rayed pectoral fin,
but is does not bring out the shape of the dorsal projection (horn) which in this

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

53

Fig. 9. The distribution of Centromochlus existimatus , from material examined.
1, Contamana. 2, Manacapuru. 3, Manaos. 4, Jatuarana. 5, Borba. 6, Villa Bella. 7, Obidos.

8, Santarem. 9, Gurupa.

54

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

specimen has a sharp angle; does not show the double fifth dorsal ray; the
anal fin is shown too far backward, reaching to well below the adipose fin;
the lateral line is shown too high on the flanks, and it looks as if the pectoral
spine is serrated on its outside.

I have considered the possibility that the differences between C. heckelii
and C. existimatus are not specific but sexual, a possibility that apppears to
be strengthened by the fact that the two have so often been collected together
(as evidenced by mixed samples in collections). The obvious way to solve
this problem was dissection of a few specimens. Unfortunately practically
all material examined was old and not very well preserved; only sample
ANSP no. 73154 (5 C. heckelii, i C. existimatus) appeared to be in a good
condition and after having obtained Dr. Bohlke’s consent, I attempted internal
sexing. This, however, was inconclusive as even in these specimens the inter¬
nal tissues had disintegrated too far for positive results. Nevertheless, I
believe to be able to distinguish between the sexes on the basis of the position
of the urogenital pore. This is placed either against the origin of the anal fin,
or well in advance of it, about half-way between the anus and the origin of
the anal fin; when considering the position in related genera, it is a reasonable
assumption that specimens of the first kind are males, of the second females.
On this character, the examined material of both species C. heckelii and C.
existimatus, consists of males and females. In addition, whereas some of the
characters which distinguish the two species are of a kind that could be sexual
(for example length of barbels), others (shape of fontanel, number of pectoral
rays, position of adipose fin) are not anywhere in the Auchenipteridae
known to show sexual dimorphism.

Centromochlus sp.

Material. — One specimen ( 9 ?), 1865/66, Lago Alexo, Brazil (S. V. R. Thayer,
Thayer Exp., MCZ no. 7634), standard length 54 mm. Three specimens (all 9 ), 19 Sep¬
tember 1909, Rio Mamore, Brazil-Bolivia (J. D. Haseman, FM no. 57804), standard
length 58 x /iy 61, 62 mm.

Characters. — D I.5, A 7 or 8, P I.7 or 1 . 8 , V 6 (i. 5), C i.is.i and rudi¬
ments, predorsal length 3.0-3.25 times in standard length, distance from tip
of snout to base of ventrals 1.8-1.85 times in standard length, maxillary
barbels reaching only a little beyond a vertical through the origin of D, pectoral
spines short, when adpressed reaching from three-quarters to four-fifths of
the distance from their base to the ventral base, and much shorter than the
predorsal length (1.2-1.3 times in this distance), the anterior edge of the
pectoral spines perfectly smooth, dorsal spine short, about four-fifths of
pectoral spines, when depressed falling just short of a vertical through the
origin of V.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

55

In all characters, including fin-formula, shape of the depression containing
the fontanels, position of adipose fin, and lateral line, the specimens are
similar to C. heckelii, but they differ by their lesser predorsal length (3.0-3.25
times in standard length, against 2.7-3.0 in C. heckelii ), short maxillary bar¬
bels and short spines.

Colour. In the specimen from Lago Alexo, there remains a colour-pattern
as described in the generic diagosis.

Discussion. — These specimens are either underdeveloped C. heckelii, or a
distinct, undescribed species. I have also considered the possibility that they
would be a cross between C. heckelii and C. existimatus, but except for the
length of the maxillary barbels, there is not a single character in which they
are intermediate. The spines are smaller than in either species, and in the
shorter predorsal length they differ also from both other species.

Genus Tatia A. de Miranda Ribeiro

Tatia A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 360 — based
on Centromochlus intermedins Steindachner and Centromochlus aulopygius Kner, type
by subsequent designation, Centromochlus intermedins Steindachner (Gosline, 1945: 10).

Generic diagnosis. — D I.4 or I.5, A 7-11, P I.4 or I.5, V (i.5), C i. 15.1
and rudiments. A genus of small catfishes with naked body, and with the
upper surface of the head (except the orbits and a small interorbital fontanel)
and the nuchal region entirely covered with bony plates; fontanel or fontanels
forming a deep oval hole between the eyes, without skull depression either
forwards or backwards; nuchal plate reaching to, and laterally beyond, the
basis of the dorsal spine; dorsal and pectoral spines strong, nearly always with
well-developed teeth and hooks, the pectoral spines thus armed on both sides,
the dorsal spine along its anterior edge only, the posterior border being
smooth; anal fin short; adipose fin small; eyes moderately large, covered with
skin; barbels of moderate length, the longest pair reaching to or just beyond
the origin of D; maximum size not over 120 mm standard length (the largest
example actually examined had a standard length of 117 mm).

Mouth terminal, not visible from below; teeth of moderate size, in a band
in each jaw, without backward projections; no teeth on vomer and palatines,
except in some large specimens of T. intermedia, which have an irregular
patch of vomerine teeth; head not very long, distance from tip of snout to
base of dorsal spine 3.25-3.7 times in standard length; horns of nuchal plate
bent outwards and downwards, sometimes bent very little, but never bent
inwards (fig. 10); humeral spine (postcleithral process) usually well-
developed, varying in length from rather less to much over half the length of
the pectoral spine (the size of this process can vary even on both sides of one

Fig. io. Heads of Tatia- species seen from above, to show differences in shape of nuchal
plates and horns, (a) T. intermedia , (b) T. perugiae , (c) T. punctata , (d) T. brunnea,
(e) T. creutzbergi , (f) T. schultzi , (g) T. concolor. 3 X.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

57

individual, and although there is certainly a tendency for it to be small in the
species of the T. aulopygia-gronp , and large in T. perugiae and T. altae, I
found it to be too variable to be useful as a specific character in the genus);
pectoral pore over middle of pectoral base; dorsal and pectoral spines com¬
paratively short, pectoral spines in all species with teeth and small hooks,
4.3-4.9 times in standard length, not nearly reaching to ventrals (the pectoral
spines are more fully described in the next paragraph); lateral line more or
less straight; anal fin with a pronounced sexual dimorphism in shape and
structure; ventrals implanted at about the middle of the body, or slightly
farther backwards; caudal fin shallowly to deeply forked, the tips of its lobes
usually more or less rounded; the caudal fin not clearly separated from the
body, hence no distinct caudal peduncle, many rudimentary rays above and
below, extending some distance forward.

Pectoral spines comparatively short, generally no more than two-thirds of
distance from tip of snout to base of dorsal spine, with a slight curve back¬
wards; on the anterior edge some 18-26 teeth, the distal ones largest, and
progressively smaller towards the body,, directed outwards, at least the
distal ones clearly so; along the posterior edge some 13-18 teeth, also largest
towards the tip of the spine, and all directed inwards; on the outer edge the
teeth continue practically to the base of the spine, along the inner edge they
usually do not come up so far (fig. 6). Exceptions: Tatia schultzi in which
the teeth along the anterior edge of the pectorals are rather smaller than in
the other species, and directed inwards, and Tatia perugiae and T. altae which
have 25-40 small teeth along the anterior edge of the pectoral spine, and only
7-10 along its posterior border (fig. 15).

The number of rays in the anal fin is not a character of diagnostic value
in the majority of species; the variation being only from seven to eleven, and
at least one species, T. creutzbergi, covers a range of from eight to eleven.
Also, the rays are not always easy to count as some rays are small and
embedded in skin, hence easily overlooked.

Most species have a distinctive and diagnostic pattern of pigmentation,
making reasonably fresh specimens easy to identify, notwithstanding a lack
of morphological characters.

Lateral line complete, but not very conspicuous, straight, with evenly spaced
pores.

Sexual dimorphism is pronounced in one character. Females have the anal
fin normal, albeit with a slightly fleshy base, and the anal opening and papilla
are situated just in advance of the fin. In males, on the other hand, the anal
fin is transformed into what one may assume is a copulatory organ; the basal
part is encased in skin, and the rays protruding from this skin form together

58

ZOOLOGISCHE VERHANDELINGEN 13 2 (1974)

a hard, more or less leaf-shaped organ, the first or second rays being thick
and long, the last rays fine and almost degenerated. At the end of the fleshy
part, hence at the base of the visible part of the first ray, is the genital
opening, widely separated from the anal opening.

There are no evident additional secondary sexual characters, and as of all
Suriname species both males and females have been available, I can state
definitely that the sexes agree in colour and pattern, as well as in the few
other specific characters I have been able to find in the genus.

Distribution. — The whole of tropical South America east of the Andes,
from Venezuela and Columbia to southern Brazil (Rio Piracicaba, Parana
drainage), but not known from the Paraguay-La Plata rivers. Not known
from Paraguay, and in Bolivia only from the Rio Guapore, the river forming
the boundaiy with Brazil.

Discussion. — The majority of species here placed in Tatia, are better
known under the generic name of Centromochlus, which has gained wide
acceptance in ichthyological literature. The name Centromochlus Kner
(1858) was based on two species, C. megalops Kner and C. aulopygius Kner.
C. megalops Kner, which by subjective judgement is a junior synonym of
C. heckelii (Filippi), has become the type of Centromochlus by subsequent
selection (Bleeker, 1862: 7).

I have examined a number of specimens of C. heckelii, including the holo-
type of C. megalops , and found that this species differs in numerous charac¬
ters from C. aulopygius (type examined) and all other species currently
placed in Centromochlus and Tatia. Although I am well aware that the genus
is the most subjective of all systematic categories, even adherents of a wide
genus-concept, to which I belong, would not regard C. heckelii as congeneric
with the other members of the group, which are morphologically extremely
uniform. Therefore the generic name Centromochlus has to be restricted to
C. heckelii and C. existimatus.

The main character on which A. de Miranda Ribeiro (1911: 353) separated
Tatia from Centromochlus was the following. u Centromochlus: papilla genital
anterior aos raios anaes; Tatia : papilla genital sobre os raios anteriores da
anal”. The genus was based on two species, T. intermedia and T. aulopygia .
The major character on which Tatia was based is one of sexual dimorphism,
as will be further discussed below, but the separation in itself was in my
opinion correct — see the supporting evidence given above.

As Tatia has never been satisfactorily defined, it is not surprising that
there has been uncertainty and confusion over which species it includes. As
an example, Gosline’s (1945) list may be mentioned, in which Tatia only
contains the two species originally placed in it by Ribeiro, and the other

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

59

species known at the time (C. oncinus, C. heckelii, C. steindachneri, C. peru¬
giae, C. gyrinus) , are retained in Centromochlus 1 ). Later Boeseman (1953)
fallowed Gosline in referring the Centromochlus aulopygius of Eigenmann
(1912) to Tatia, and described a new species as Centromochlus creutzbergi .
Other species described in recent years, have been placed in Centromochlus
without reference to Tatia (cf. Fowler, 1945a; Rossel, 1962). P. Ribeiro
(1962b: 9, 10) listed one of the species twice, on one page as Centromochlus
aulopygius, on the next page as Tatia aulopygia.

The genus Tatia as here defined consists of a group of small fishes of
extremely uniform morphology. At present it appears possible to distinguish
fourteen species in it, five of which are known to occur in Suriname. Of the
five Suriname species, three are described here for the first time and another
three new species were found in collections from other countries. This sug¬
gests that the genus is by no means well known, and that many more species
remain to be discovered in various parts of its range.

The time is not ripe, in my opinion, for the application of ternary nomen¬
clature in this still very insufficiently known genus, but I regard it as likely
that some of these forms can be grouped as follows:

T. aulopygia aulopygia

T. aulopygia intermedia

T. aulopygia neivai

T . aulopygia galaxias

T. perugiae perugiae

T. perugiae altae

On the basis of present knowledge, no further grouping together of forms
appears possible. The co-occurrence of four species of this very uniform
genus in the Suriname Rivier basin is a warning against oversimplication.

Tatia aulopygia (Kner) (figs. 11, 13, pi. 2)

Cent\romochlns] aulopygius Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw.
Cl., 26 (1857) : 432, pi. 8 fig. 25 — Rio Guapore.

Centromochlus aulopygius : Steindachner, 1876, Sitzb. Akad. Wiss. Wien, Mathem.-
Naturw. Cl., 74 (1) : 664, 665 (no locality); Pearson, 1937, Proc. Calif. Acad. Sci., (4)
23: no (Mamore basin).

1) Arius oncina R. H. Schomburgk (1841) included in Centromochlus bij Gosline, is
a doubtful species. R. Schomburgk (1848: 618) has stated concerning his brother's book
that: “...die dem Buche beigefiigten Abbildungen nur in wenigen Fallen auf Treue
Anspruch machen konnen”. The peculiar serrated scapular spine, the shape of the
caudal fin, and its maximum length of ten inches, are all unlike any known species of
Tatia .

6o

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

6l

Tatia aulopygia ; (pt.) A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro,
16: 361 (Guapore); (pt.) Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 10
(Guapore).

Centromochlns intermedins; (pt.) Fisher, 1917, Ann. Carnegie Mus., 11: 422 (Maciel,
Rio Guapore); Pearson, 1937, Proc. Calif. Acad. Sci., (4) 23: no (Marnore basin).

Material. — One specimen, “1817”, Rio Guapore (Natterer, MV no. 47332), total
length 52, standard length 41 mm, syntype of the species 1 ). Three specimens, 3 August
1909, Maciel, Rio Guapore (J. D. Haseman, FM no. 58015), standard length 23, 35,
41^2 mm. Nine specimens, 13 May 1968, Suia Missu, Mato Grosso (R. H. McConnell,
BM no. 1971.7.29: 6-12, and two specimens RMNH no. 26491), standard length 19-31 mm.

Characters. — Of medium size (standard length 19-14J/2 mm); nuchal plate
with horns well flung outwards, pointed; hind-border of adpressed anal fin
opposite hind-border of adipose fin, or not reaching that far; usual number
of rays in the anal fin 10; colours and colour-pattern diagnostic.

Colour. Head and body entirely dark grey-brown; the head and predorsal
part of the body darkest, blackish grey; under parts from lower chin to anus
unpigmented, white; body with numerous small whitish, short longitudinal
stripes, which are distinct only when viewed at close quarters; seen from
some distance the colour is even blackish-brown; dorsal fin with the skin
covering the spine, and the basal area dark brown, remainder hyaline; pec¬
torals with scattered pigment over the spines, further hyaline; anal with a
brownish basal half, the distal part hyaline; ventrals hyaline; caudal base
pigmented as the body, caudal fin pigmented with the same pattern, but much
more lightly, the pigment being most conspicuous in the upper part of the
upper lobe, and in the lower part of the lower lobe; adipose fin with irregular
brownish pigment.

Distribution (fig. 13). — Probably widely distributed in the centre of
South America, but at present only known from the Rio Guapore and the
Suia Missu.

Discussion. — In recent years the name Tatia aulopygia or Centromochlus
aulopygius has been universally applied to the species with a pattern of large
round and elliptical white (unpigmented) spots on an earth brown or grey
background. Being interested to know if the type specimens would still show
remnants of this characteristic, and for the species diagnostic, pattern, I
applied for a loan of a syntype, a request that was most kindly granted. The
specimen did not appear to have had the pattern with large white spots but
is more or less evenly pigmented, with the belly pigmentless. The pigmenta¬
tion of the chin differs conspicuously from that of “ aulopygia ” auct. by

1) There is no original label and the year 1817 cannot be correct, as Natterer’s visit
to the Rio Guapore took place in 1828/29 (cf. Pelzeln, 1868-1871: Itinerarium).

62

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

consisting of more or less evenly spaced large melanophores, whereas in the
latter species there is a distinct pattern: usually a large pigmented V sepa¬
rated from the pigment along the edge of the mandible by an almost pigment¬
less area; this pigmented area varies in extent, and is sometimes entirely
absent (pi. 2). Whereas I can imagine that in a specimen as old as the type
of T. aulopygia pigment has worn off or been lost in some other way (this
evidently happens in old material), it is unlikely that pigment could appear
where originally it was not present. Notwithstanding the close morphological
resemblance (shared by most other species of the genus), I believe therefore
that T. aulopygia is a species (or subspecies) different from the one that over
the past sixty years has been called by that name, and that the latter must bear
the name T. intermedia.

Besides the evidence presented above, it should be noted that Kner (1858),
who described T. aulopygia less than thirty years after the specimen had been
collected, did not mention a distinctive pattern. Some specimens of T. inter¬
media examined by me, collected well over thirty years ago, even several taken
over a century ago, still show its diagnostic pattern.

In this connection it is significant that the only specimens of Tatia that have
since been collected in the Rio Guapore, at a time that specimens with a grey-
and-white pattern were generally called C. aulopygius, were listed not under
that name, but as C. intermedins (cf. Fisher, 1917). I have examined Fisher's
specimens; they are dull brown, with small longitudinal pale streaks and the
largest has its chin pigmented just as the type of T. aulopygia. It is likely
therefore that these specimens actually represent T. aulopygia , although cer¬
tain identification of this old material is no longer possible.

I was very uncertain about the characters of this species until I received Mrs.
McConnell's sample from the Suia Missu river: these specimens have, as
described in the paragraph “Colour” which is based on them, a pattern of
small pale streaks, different from all other known species, but similar to that
of Fisher’s Guapore specimens. Of course the distance separating the Suia
Missu (a tributary of the Xingu) from the Guapore (the upper course of
the Madeira) is considerable, and geographically the identity of the popula¬
tions inhabiting these two rivers does not suggest itself as obvious, but the
agreement in colour pattern between specimens from the two rivers, which is
supported by such morphological evidence as exists, causes me to assign them
to one species with some confidence.

A specimen of uncertain provenance which died in the aquarium of “Artis”,
Amsterdam, on 12 November 1970, and was forwarded to me a few days
later, belongs either to this species, or to a form extremely close to it. It
differs from the material discussed by its larger size (standard length

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

63

64 mm), and this may account for some or all of the other differences: body
black, without any trace of paler markings, except for the ventral region,
from an arc bent forwards between the bases of the pectorals to just before
the ventrals, which is pigmentless, white; caudal fin rather deeply forked,
with the upper lobe the longer; basal third of ventral fins pigmented, black;
posterior border of adipose fin reaching farther backwards than posterior
border of anal fin; there appears to be a suggestion of vomerine teeth. The
spcimen is now ZMA no. 110698.

Tatia intermedia (Steindachner) (figs. 5, 6, 10a, 12, 13, pi. 2)

Centromochlus intermedins Steindachner, 1876, Sitzb. Akad. Wiss. Wien, Mathem.-
Naturw. CL, 74 (1): 664, footnote 1 — Marabitanos, Para.

Centromochlus dunni Fowler, 1945, Proc. Acad. Nat. Sci. Philad., 97: in — Morelia,
Rio Caqueta drainage, Colombia.

Centromochlus aulopygius; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 198 (Essequibo);
(pt.) Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34 (Essequibo); (pt.)
Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 395 (Essequibo) ; Eigenmann,
1912, Mem. Carnegie Mus., 5: 197, pi. XX fig. 1 (Creek below Potaro Landing; Wis-
mar) ; Arnold & Ahl, 1936, Fremdl. Siisswasserfische: 255, fig. (das ostliche tropische
Siidamerika von Guyana bis zum Stromgebiet des Amazone); Puyo, 1949, Poiss. Guyane
Fr.: 97 (les criques de la region de l’Approuage) ; Sterba, 1959, Siisswasserfische aus
aller Welt: 267, fig. 539 (Ostliches Siidamerika nordlich des Amazonas) ; P. de Miranda
Ribeiro, 1962, Publ. Avuls. Mus. Rio de Janeiro, 45: 9 (Rio Araguaya, Aruarua (antiga
Leopoldina), Estado de Goiaz) 1 ).

Centromochlus Perugiae; Vaillant, 1899, Bull. Mus. Hist. Nat. Paris, 5: 155 (le haut
du fleuve Carsevenne: la riviere Lunier ou la riviere Carnot) ; Vaillant, 1900, Nouv.
Arch. Mus. Paris, (4) 2: 124, 127 (la riviere Carnot).

Centromochlus intermedins ; Steindachner, 1882, Denkschr. Akad. Wiss. Wien,
Mathem.-Naturw. Cl., 44 (1) : 4 (Hyutahy und Jatuarana) ; Eigenmann & Eigenmann,
1888, Proc. Calif. Acad. Sci., (2) 1: 156 (Tajapuru, Teffe, Jatuarana, Iga, Jutahy, Lago
Alexo) ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1: 269 (Tajapuru,
Teffe, Jatuarana, Iga, Jutahy, Lago Alexo) ; Eigenmann & Eigenmann, 1891, Proc. U.S.
Nat. Mus., 14: 34 (Amazon, Solimoens and tributaries); (pt.) Fisher, 19x7, Ann. Car¬
negie Mus., 11: 422 (Rio Tapajos at Santarem).

Tatia intermedia ; A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16:
360 (Maribitanos, Jutahy, Jutuarana, Tajapuru, Iga, Lago Aleixo) ; Gosline, 1945, Bol.
Mus. Nac. Rio de Janeiro, Zool., 33: 10 (Amazonas; Solimoes e tributarios).

Tatia aulopygia ; (pt.) Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 10
(Essequibo); (pt.) Fowler, 1951, Arq. Zool. S. Paulo, 6: 470 (Guiana); P. de Miranda
Ribeiro, 1962, Publ. Avuls. Mus. Nac. Rio de Janeiro, 45: 10 (Amazonas); Lowe
(McConnell), 1964, J. Linn. Soc. Lond., Zool., 45: 116, 141 (Simoni Creek, Rupununi
District); McConnell, 1967, Timehri, 43: 69, 70 (Rupununi) ; P. de Miranda Ribeiro,
1968, Bol. Mus. Nac. Rio de Janeiro, Zool., 263: 10, fig. IX (no locality = Amazonas).

Centromochlus dunni; Fowler, 1951, Arq. Zool. S. Paulo, 6: 462 (Alto Amazonas,
Colombia); Rossel, 1962, Senckenb. Biol., 43: 20 (no locality); Liiling, 1963, Beitr. Neotr.
Fauna, 3: 50, fig. 15 (Quisto Cocha, Iquitos).

1) Hoedeman's (1968: 150) coloured figure, captioned Centromochlus aulopygius , does
not resemble either Tatia aulopygia or Tatia intermedia; I have no idea what it is.

64

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Material. — Three specimens ( 5 , 2 9 ), before 1864, Essequibo (Ehrhardt, BM
no. 64.1.21.19-21), standard length 66 ( $ ), 86, 89 mm. One specimen, Jutahy (no collector,
presumably Thayer Exp., MCZ no. 8171), standard length 22^4 mm. One specimen, I^a
(no collector, presumably Thayer Exp., MCZ no. 7344), standard length 27 }4 mm. One
specimen, ca. 1866, Lago Alexo, Brazil (S. V. R. Thayer, Thayer Exp., MCZ no. 8176),
standard length 31 mm. One specimen ( 9 ), 1866, Tajapura (L. Agassiz, Thayer Exp.,
MCZ no. 7333), standard length 117 mm. One specimen ( $ ), 1866, Teffe (L. Agassiz,
Thayer Exp., MCZ no. 7375), standard length 97 mm. One specimen, same data (L.
Agassiz, Thayer Exp., MCZ no. 8154), standard length 26^4 mm. One specimen, before
1874, Para (MV no. 47335), standard length 73 mm, syntype of the species. One specimen,
before 1875, Hyutahy (MV no. 47356), standard length 42 mm. One specimen ( 9 ), May-
September 1898, Carnot, tributary of the Carsevenne, now Calgoene, State of Amapa,
Brazil (F. Geay, MNP no. 99-99), standard length 47 mm. One specimen ( $ ), no date
or collector, but undoubtedly from Eigenmann's collection, 1908/1909, Wismar, British
Guiana (CAS no. 24798), standard length 4754 mm. One specimen (S), 1912, Amazonas
(Steindachner, MV no. 47334 Z), standard length 64 mm. One specimen ( 9 ), not dated,
ca. 1913, Serra de Melle, Rio Surumu (Haseman, MV no. 47335), standard length 41 mm.
One specimen ( 9 ), August 1924, Santarem (C. Ternetz, CAS no. 6678), standard length
74 mm. One specimen (£), not dated, creek below Potaro Landing, British Guiana
(Shideler, FM no. 53261), standard length 49 mm. One specimen ( 9 ?), April 1937,
Rockstone, Essequibo River (E. R. Blake, FM no. 42858), standard length 36 mm. One
specimen ($), 1945, Morelia, Caqueta drainage, Colombia (Kjell von Sneidem, ANSP
no. 71706), standard length 52 mm, paratype of Centromochlus durmi. 16 specimens,
4 May 1961, Simoni Tacuba, Rupununi (R. H. Lowe, no. 655), standard length 32-78 mm.
Four specimens (4 9 ), July 1965, Igarape, Paracuri, Belem near Icoaraci, Para, Brazil
(N. Menezes, MCZ no. 46130, one specimen now RMNH no. 26492), standard length
36, 41, 41, 41 mm. One specimen, July 1965, Rio Arari, Cachoeira do Arari, Ilhas de
Marajo, Para, Brazil (N. Menezes, MCZ no. 46032), standard length 20 mm. 11 speci¬
mens (8 $, 3 9 ), 23 January 1966, Sipilawini (Mees, RMNH no. 26199), standard
length 38^4-50 mm. One specimen ( 9 ), 26 January 1966, Sipaliwini (Mees, RMNH
no. 26200), standard length 45 mm. Two specimens ( $, 9), 2 February 1966, Sipaliwini
(Mees, RMNH no. 26195), standard length 50, 54 mm. One specimen (9 ), 30 January
1967, Gran Mau Kreek, tributary of the Gran Rio near Dombaai (Nijssen, ZMA
no. 105523), standard length 38J4 mm. One specimen ( 9 ), 31 January 1967, tributary
of Gran Rio 4 km N. of Awaradam (Nijssen, ZMA no. 105524), standard length 35 mm.
Eight specimens (6 $, 2 9 ), 20 March 1967, Parwapa Kreek, tributary of the Suriname
Rivier near Botopassie (Nijssen, ZMA no. 105791), standard length 50-74 mm. Three
specimens (2 $, 9 ), 5 April 1967, tributary of the Nickerie Rivier ca. 12 km S. of
Stondansie Vallen (Nijssen, ZMA no. 105831), standard length 34, 54 ( 9 ), 60 mm.
One specimen, 10 August 1968, creek near aerodrome Kaysergebergte (M. S. Hoogmoed,
RMNH no. 26201), standard length 27 mm. Two specimens (9 and ?), undated, Rio
Araguaya, Aruana (antiga Leopoldina), Estado de Goiaz (C. Griem (Aquario Rio),
MRJ no. 5896), standard length 28, 65 mm.

Characters. — Size large (standard length 20-1x7 nun)} nuchal plate with
blunt horns, moderately bent outwards; hind-border of anal fin opposite
hind-border of adipose fin; usual number of rays in the anal fin 10 (iii-7);
male anal fin pointed, with its third and fourth rays longest; colour and
colour pattern diagnostic.

The male anal fin develops at a standard length of 35 to 40 mm.

Colour. Ground colour dark earth brown or grey-brown, with many large

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

65

5

66

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

roundish to elliptical (with a longitudinal direction) white dots; in large speci¬
mens these dots tend to be relatively smaller. Under parts from chin to anus
white, not pigmented, except along the edge of the lower jaw, and often a
V-shaped band of pigment across the lower jaw; the development of this
band is variable, and it may be irregular or practically absent. Our preserved
specimens have retained their colour well, but in old material, collected from
sixty to over a hundred years ago, the ground colour has changed to light
brown, and I have felt it justified to include in the species some old specimens
in which no pattern or original colour is now visible.

Distribution (fig. 13). — This is on the basis of present evidence certainly
by far the commonest and probably the most widely distributed member of
its genus. It ranges throughout the northern and western part of the Amazon
basin, from Colombia and Peru to the mouth of the Amazon (Para =
Belem), and into the Guianas: Amapa, French Guiana (Puyo, 1949), Suri¬
name and British Guiana. Not known from Venezuela (see discussion). It is
unlikely that the specimens from Sao Paulo, listed by A. de Miranda Ribeiro
1918b : 734) and P. de Miranda Ribeiro (1962b: 10) under the name Tatia
intermedia were correctly identified; a discussion of these specimens is given
under T. neivai .

Discussion. — From the syntype from Para, the colour and pattern had
entirely rubbed off, and could therefore not assist in its proper allocation. A
specimen from Hyutahy ( = Jutahy = Jutai), however, which is non-typical
but belongs to the material identified and described by Steindachner (1882)
very soon after the establishment of the species, does still show the charac¬
teristic colour pattern. See also the description of colour and pattern given
by Steindachner (l.c.). Further support for the identification of T. intermedia
with the spotted species is that several collectors have obtained it, and no other
member of the genus, at Para.

Most specimens from the Thayer collection, although now over a century
old, still show at least traces of the diagnostic colour pattern. It is evident in
the very large specimen of 1x7 mm. This is surprising inasmuch as Eigen-
mann & Eigenmann (1890), whose description was based on the MCZ-
material, stated: “Young with short whitish streaks along the sides”, thus
implying that these streaks do not occur in large specimens.

In two of the examined specimens I found vomerine teeth present: a very
small patch in a female from Essequibo (standard length 89 mm; BM no.
64.1.21. 19-21), a quite distinctive patch of well-developed teeth in a female
from Tajapura (standard length 117 mm, MCZ no. 7333). I have not found
palatal teeth in any other species of the genus, but no specimens of other
species of a similar size were available — likely no other species reaches such

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

6 7

Fig. 13. The distribution of Tatia intermedia (dots), T. aulopygia (triangles), T. neivai
(star), and T. galaxias (circles).

68

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

a large size. On the other hand, the presence of these teeth may not be exclu¬
sively a matter of size, as the second largest specimen of the series examined
(( 3 , standard length 97 mm, MCZ no. 7375) has none.

In the description of C. dunni, Fowler (1945a) mentioned as differential
character from “C. aulopygius" (= Tatia intermedia ), the following: “The
photographic figure published by Eigenmann as Centromochlus aulopygius
Kner is from a specimen 65 mm long showing a different color pattern, with
much smaller white spots short in size, besides the anal fin of different shape”.
The difference in shape of the anal fin is of no consequence, being sexual:
Eigenmann figured a female, whereas type and paratypes of C. dunni are males.
This reduces the difference between T. intermedia and C. dunni to one of
size and intensity of the markings. I have examined the paratype of C. dunni
and found that, as stated by Fowler, the white markings are larger and less
well-defined than in fresh material from Suriname. Originally this difference
appeared to me sufficient to keep C. dunni provisionally separate from T.
intermedia, but on reconsideration I have decided, albeit with some hesitation,
that the difference is probably within the limits of individual variation. There
is, within the diagnostic basic pattern, a lot of individual variation in shape
and extent of the white markings, and amongst others a specimen from the
Carcevenne, Amapa, has also large white markings, similar to C. dunni ,
although the geographical distance separating the two populations is enormous.

The specimen from the Carcevenne, just mentioned, is the one listed by
Vaillant (1899, 1900) under the name Centromochlus Perugiae. Examination
revealed that it belongs to T. intermedia, but its spots are larger and more
horizontally elongated than in our specimens from Suriname, and the
posterior border of the adipose fin is definitely farther backwards than the
posterior border of the anal fin.

I have not examined the fish recorded by Liiling (1963) under the name
Centromochlus dunni, but his figure shows a typical individual of T. inter¬
media , with smallish white spots.

The largest specimen of sample MRJ no. 5896 looks like a typical example
of T. intermedia , although on geographical grounds I would have expected
T. aulopygia in the Rio Araguaya. Note that the specimen was received from
a dealer (Aquario Rio), so that its provenance may be not above suspicion;
for that reason I have not indicated Aruana on the distributional map. About
the second specimen of this sample I am in doubt, it may belong to a differ¬
ent species.

The specimens of sample MCZ no. 46130 from near Belem have the white
dots remarkably well-defined, each with a very narrow dark edge, which is
absent in specimens collected in Suriname in the same year. Also, the white

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

69

dots do not extend to the back as they do in other material. For the moment
I assume that this difference falls within the range of individual variation.

Venezuela has been included in the range of u Centromochlus aulopygius”
(= Tatia intermedia ) on the basis of a specimen recorded by Pellegrin
(1899). The specimen does not belong to this species and will be discussed
under Tatia galaxias .

I have examined the specimen listed as Centromochlus intermedins by
Marlier (1968: 53): <$, 28 January 1964, Rio Prato da Eva, Amazonas
(G. Marlier, KBIN no. 23156), standard length 30 mm. The specimen does
not appear to belong to T. intermedia ; it is (in a preserved condition), light
brown, with many (about 18 on each side) narrow longitudinal white bands
over the whole length of the body; the caudal fin is forked with, apparently,
acute tips; the adipose fin is comparatively well-developed, and reaches
farther backwards than the anal fin. In shape of the nuchal horns and in
number of anal rays (10) it agrees with T. intermedia . Probably the speci¬
men represents another undescribed species, mainly characterized by colour-
pattern, but as the colours are faded, the caudal fin is somewhat frayed, and
both pectoral spines are damaged, there remains an element of doubt which
causes that I prefer to leave the specimen unnamed.

Tatia perugiae (Steindachner) (figs. 10b, 14, 15)

Centromochlus Perugiae Steindachner, 1883, Denkschr. Akad. Wiss. Wien, Mathem.-
Naturw. Cl., 46 (1) : 29, pi. VII fig. 2, 2a — Canelos, Ecuador.

Centromochlus perugiae; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2)
1: 137 (name only); Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1: 270
(Canelos); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34 (Canelos);
Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 395 (Canelos) ; Eigenmann &
Allen, 1942, Fish. W. South America: 118 (Canelos); Gosline, 1945, Bol. Mus. Nac.
Rio de Janeiro, Zool., 33: 10 (Canelos) ; Fowler, 1951, Arq. Zool. S. Paulo, 6: 464
(Alto Amazonas, Equador); Rossel, 1962, Senckenb. Biol., 43: 27-30 (discussion of
types) ; Ovchynnyk, 1968, Zool. Anz., 181: 243 (Ecuador: Canelos on Rio Bobonaza,
tributary of Rio Pastaza, Amazon basin, Prov. Napo-Pastaza) ; P. de Miranda Ribeiro,
1968, Bol. Mus. Nac. Rio de Janeiro, Zool., 263: Est. VIII fig. 2 (no locality, figure
only, after Fowler).

Centromochlas; Lundberg & Baskin, 1969, Amer. Mus. Novit., 2398: 6, 26 (no
locality).

Material. — One specimen ($), Canelos, Ecuador (received from Steindachner in
1882, MV no. 47358), standard length 39 mm, syntype of the species. One specimen
( $ im.), November 1938, Sarayacu, Ecuador (W. Clarke-Maclnture, UMMZ no. 173420),
standard length 31 mm. Two specimens (both $), summer 1967, Rio Aguarico near
Santa Cecelia (W. G. Saul, UMMZ on loan from Kansas Mus., no number), standard
length 38, 40 mm.

Characters. — Of medium size (standard length 31-40 mm, the larger sizes
of to 48 and 54 mm given by Fowler and Steindachner refer to total length);

70

Z 00 L 0 GISCHE VERHANDELINGEN I32 (1974)

MEES, AUCHENIPTERIDAE AND PIMELODIDAE Jl

nuchal plate broad, not constricted in advance of D, and its horns also broad,
hardly bent outwards; pectoral spine different from that of all other species
except T . altae, with along its anterior edge 25-40 small teeth, and only 7-10
along its posterior border (fig. 15); hind-border of the adpressed anal fin
reaching to beyond the hind-border of the adipose fin; number of rays in
the anal fin 7 or 8; male anal fin strongly modified, slender and pointed, its
second ray by far the longest and strongest; caudal fin forked with apparently
almost acute tips to the lobes; colour and colour pattern diagnostic.

Colours. There is a very distinctive pattern of large blackish-brown blotches
on a creamy-white unpigmented background; these dots are most pronounced
on the back, less so on the dorsal surface of head and nape, much vaguer on
the sides, and absent from the lower flanks and underparts, which are unpig¬
mented. Even under very low magnification it can be seen that the blotches
are made up of groups of melanophores; where these are closer together, the
dots become darker. Although Steindachner’s syntype has been in preservative
for close on a century, it fully retains its characteristic colour-pattern.

Distribution. — Known from Canelos and Sarayacu, Rio Bobonaza,
Maranon drainage, Ecuador, and Santa Cecelia, Rio Aguarico, Rio Napo
drainage, Ecuador.

Fig. 15. Right hand pectoral spine of Tatia perugiae. 6 X. Compare with fig. 6

(T. intermedia ).

Tatia neivai (R. von Ihering) (figs. 13, 16)

Glanidium neivai R. von Ihering, 1930, Arch. Inst. Biol. S. Paulo, 3: 99, pi. 13 fig. 1 —
Piracicaba, rio Piracicaba, Est. de S. Paulo.

Tatia intermedia; A. de Miranda Ribeiro, 1918, Rev. Mus. Paul., 10: 734 (Piquete,
S. Paulo; Piracicaba); (pt.) Fowler, 1951, Arq. Zool. S. Paulo, 6: 470 (Sao Paulo);
P. de Miranda Ribeiro, 1962, Publ. Avuls. Mus. Nac. Rio de Janeiro, 45: 10 (Piquete,
Estado de Sao Paulo).

Glanidium neivai; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 11 (Sao
Paulo); P. de Miranda Ribeiro, 1962, Bol. Mus. Nac. Rio de Janeiro, Zool., 240: 3
(sudeste do Brasil).

Material. — One specimen ($), undated, Rio Piracicaba near Piracicaba, State of Sao
Paulo (Escola Superior de Agricultura Luiz de Queiroz, MZSP), standard length
57 mm.

Characters. — Size large (standard length 57-58 mm); nuchal plate with
horns well flung outwards, pointed, as in T. aulopygia; hind-border of

72

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

adpressed anal fin opposite hind-border of adipose fin, or not quite reaching
that far; number of rays in the anal fin io; caudal fin forked with rounded
lobes, the upper lobe somewhat larger than the lower lobe; colour-pattern
diagnostic.

Colour of specimen preserved for an uncertain time. Head and body dull
brown, except for the underparts from lower chin to anus, which are unpig-
mented whitish; body with not very numerous and rather small whitish dots,
the ones situated most dorsally almost round, those on the flanks horizontally
elongated; dorsal fin with the skin covering the spine, and the basal area dull
brown, the first and second rays with their basal halves dull brown, remainder
of the fin hyaline; anal base dull brown as the body, the small, modified fin
hyaline; pectorals with scattered pigment on the spines, and a few dispersed
melanophores elsewhere, further hyaline; ventrals hyaline; caudal fin with
alternating irregular dull brown and whitish vertical bands of equal width,
five of each; adipose fin with dull brown base and centre, its edges hyaline.

This description was based on the single available specimen, a male with
fully modified anal fin, but I have also compared R. von Tiering’s (1930)
description and figure of the holotype, a female of 58 mm standard length.
The present location of the type is unknown and it may be lost.

Distribution (fig. 13). — With certainty only known from the type-locality,
the Rio Piracicaba, a tributary of the Rio Tiete, Parana drainage. State of
Sao Paulo. Specimens from Piquete listed as Tatia intermedia by A. de Mi¬
randa Ribeiro (1918b) and P. de Miranda Ribeiro (1962b) may belong here,
as tentatively suggested by their inclusion in the bibliography above, but this
is doubtful, for, although Piquete is geographically not far from Piracicaba,
it is on the Rio Parahyba, which flows to the north-east, remote from the
Parana. I have examined two of the specimens from Piquete (MRJ no. 940),
standard length 36 and 38 mm, but they are in such a poor condition that a
certain identification appears impossible. Such morphological characters as
they have, do not contradict the suggested identification. The specimens
appear to have been stained, and are more or less transparent; they seem also
to have been dry; bones and spines are decalcified and soft, and have lost
much of their structure.

Discussion. — Although it was originally described in Glanidium , this
ispedes definitely belongs to the genus Tatia, as proved by the thin skin of
the head, through which the bone is clearly visible, the shape of the horns of
the nuchal plate, the dorsal and pectoral spines which have teeth along the
whole length of their anterior edges, the four-rayed pectoral fins, and the
low number of rays in the anal fin. The dorsal spine has 13 teeth along its
anterior edge; the right pectoral spine (the other one is broken near its base),

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

73

74

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

has 18 teeth along its anterior edge, n larger ones along its posterior edge :
counts which are normal for the genus.

This species agrees morphologically with T. aulopygia, T. intermedia and
T. galaxias, but differs from each of these in colour-pattern as follows: from
T. aulopygia by its distinct and not numerous white spots, from T. inter¬
media by its much smaller and more sparse white spots, from T. galaxias by
much sparser spots and (probably also in fresh material) less dark body.

Tatia gyrina (Eigenmann & Allen)

Centromochlus gyrinus Eigenmann & Allen, 1942, Fish. W. South America: 118,
pi. 5 fig. 4 — brook near Rio Itaya, Iquitos.

Centromochlus gyrinus; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 10
(Iquitos); Fowler, 1945, Los Peces del Peru: 62 (Iquitos); Fowler, 1945, Proc. Acad.
Nat. Sci. Philad., 97: 112 (no locality); Fowler, 1951, Arq. Zool. S. Paulo, 6: 462,
fig. 491 (Alto Amazonas, Peru).

Material. — None. Some notes on the holotype, additional to the published ones, were
taken at my request by Dr. Eschmeyer, and will be given below.

Characters. — Probably of medium size (standard length of only known
specimen ca. 42 mm); nuchal shield broad, hardly constricted, its horns small
and hardly bent outwards; hind-border of anal fin clearly in advance of hind-
border of adipose fin; anal fin with 9 rays; colour pattern diagnostic.

On illustrations (Eigenmann & Alien's photograph, and Fowler's figure
which was copied from it), it looks as if the adipose fin is very long, originat¬
ing not far behind the dorsal fin, anteriorly low and increasing in height
backwards; in the description the adipose fin is referred to as “well-marked,
merging forward"; this would be unlike any other member of the genus, but
information received from Dr. Eschmeyer is that the adipose fin is in fact
short, and originates about opposite the end of the anal base.

In the original description the size of the eye is given as a character of
some importance: eye 3.3 in head, 2.5 in interorbital, 5 in snout (an obvious
misprint for 0.5), and: “near to C. intermedins, and differing in ... larger
eye However, on the illustrations the eye does not look large but, on the
contrary, smaller than in other species. Measuring from the photograph, I
find that the eye is fully six times in the head. Dr. Eschmeyer has confirmed
that the size of the eye has been illustrated correctly. The reason why I have
not used the size of the eye in the study of the genus, is in the first place that
there did not appear to be much difference between the species, and secondly
that it is difficult to measure. Not only is the eye entirely covered by skin,
but also it is imbedded in transparant connective tissue and fluid, which may
or may not be included when measuring it. In the figures illustrating mem-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

75

bers of the genus, incidentally the larger outline of the eyes has been shown.
Even taking the smaller outline, however, the eye of T. intermedia is only
about four times in the head, hence definitely larger, not smaller than the
eye of T. gyrina.

The fin-ray numbers given in the original description are D 1 . 6 , A 12.
For both dorsal and anal fin these numbers are higher than in any other
species of the genus; Dr. Eschmeyer wrote, however: “I believe the dorsal
rays are I + 4, and the anal rays appear to be about 9 and almost certainly
not 12”. An X-ray photograph received from Dr. Eschmeyer shows clearly
D I.4, A 9 (perhaps 10).

Distribution. — Known from the type-locality only.

Discussion. — With the corrected information on fin-ray numbers and
shape of the adipose fin, it will be clear that T. gyrina is a normal represen¬
tative of its genus. A direct comparison with its congeners remains to be made.

Tatia altae (Fowler) (fig. 17)

Centromochlus altae Fowler, 1945, Proc. Acad. Nat. Sci. Philad., 97: 109 — Morelia,
Rio Caqueta drainage, Colombia.

Centromochlus altae ; Fowler, 1951, Arq. Zool. S. Paulo, 6: 462 (Alto Amazonas,
Colombia); Rossel, 1962, Senckenb. Biol., 43: 30 (no locality).

Material. — Two specimens ( $, 9), undated, Morelia, Rio Caqueta drainage, Colom¬
bia (Kjell von Sneidern, ANSP from nos. 71701-71704), standard length 30 (9),
35 ($ ) mm, paratypes of the species.

Characters. — Morphologically in every respect similar to P. perugiae,
except perhaps body somewhat deeper, but colour pattern distinct and diag¬
nostic.

Colours. Dorsal surface and flanks blackish-brown, with narrow white
reticulations.

Distribution. — Known from the type-locality only.

Discussion. — In every respect: the broad nuchal plate with horns hardly
bent outwards, the pectoral spines, shape and length of the male anal fin,
low number of rays (7 or 8) in the anal fin, and shape of the caudal fin, this
form agrees with T. perugiae, but its colour pattern is different although
clearly related. Compared with T . perugiae the dark dots have increased in
size and intensity, and are separated only by narrow white lines, giving the
fish a reticulated appearance. Whereas in T. perugiae there is near the base
of each caudal lobe an ill-defined roundish grey dot, in T. altae these dots are
well marked and tapering backwards to very near the tail tips.

There is no doubt in my mind that T. altae is very closely related to T .
perugiae, of which in future it will probably have to be regarded as a sub-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

77

species, but at the same time the differences in colour-pattern are so marked
that T. altae is certainly a valid form.

Fowler (1954a) described this species as having seven ventral rays, a
number different from all other members of the genus, but both paratypes
examined by me have six (i.5) ventral rays.

Tatia creutzbergi (Boeseman) (figs, ioe, 18, 19)

Centromochlus creutzbergi Boeseman, 1953, Zool. Meded., 32: 7, fig. ic — Djaikreek.

Centromochlus aulopygius ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) 1: 157 (Cudajas); Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1:
270 (Cudajas); (pt.) Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34
(Cudajas).

Centromochlus creutzbergi ; Hoedeman, 1957, Beaufortia, 6: 151, fig. 8 (Coropina
creek, near Republiek); Rossel, 1962, Senckenb. Biol., 43: 30 (no locality); Hoedeman,
1968, Elseviers Aquariumvissen Encycl., 1: 148 and fig. (Suriname).

Material. — One specimen ( 9 ), 1865/66, Cudajas (Thayer Exp., MCZ no. 8182),
standard length 32*/2 mm. One specimen ( 9 ), 8 October 1948, Djaikreek (D. C. Geijskes
and P. Creutzberg, RMNH no. 19440), standard length 26 mm, holotype of the species.
One specimen ( 9 ), 20 October 1956, Coropina Kreek near Republiek (J. van der Kamp,
ZMA no. 102006), standard length 30^2 mm. One specimen ($), 27 February 1967,
tributary of Kleine Saramacca Rivier, ca. 11 km E.S.E. of its mouth (Nijssen, ZMA
no. 105657), standard length 33 mm. Four specimens (all 9 ), 5 April 1967, tributary of
Nickerie Rivier, ca. 12 km W.S.W. of Stondansie (Nijssen, ZMA no. 105832), standard
length 27, 29, 31, 32 mm. 49 specimens (£, 9 ), 14 December 1966, Carolina Kreek
(Nijssen, ZMA no. 105859), standard length 21-33 mm.

Characters. — Size small (standard length 21-33 mm), nuchal shield broad
and hardly constricted laterally, with horns which are blunt and hardly curved
outwards; teeth along anterior edge of dorsal spine only weakly developed;
hind-border of anal fin clearfy in advance of hind-border of adipose fin;
number of rays in the anal fin 8 to 11; caudal fin forked shallowly; body
relatively heavy; colour and colour pattern diagnostic.

Colours. Upper surface of head and back dull brown, sides longitudinally
maibled brown and white, a broad blackish longitudinal band on the sides;
chin, underparts of opercles and isthmus marbled brown and white, breast
down to anus white (devoid of pigment). Dorsal fin dull blackish brown,
distally hyaline, caudal fin hyaline lightly spotted with brown, above and
below a fairly distinct brown streak. The most conspicuous character of this
species is the dark lateral band.

Note. Of the largest lot of 49 specimens, at least 20 are males. They show
that the male character develops at a standard length of ca. 25 mm. Smaller
specimens cannot be certainly sexed on external morphology.

I do not know why Hoedeman (1957) regarded ZMA no. 102006 as a
male: hi® opinion cannot have been based on examination of its gonads, as

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

79

Fig. 19. The distribution of Tatia creutzbergi, from material examined. 1, Maratakka (recently collected material, not
mentioned in the text). 2, Nickerie Rivier near Stondansie. 3, Kleine Saramacca. 4 Coropina Kreek. 5, Carolina Kreek.

6, Djaikreek. 7, Cudajas.

8o

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

the specimen has not been dissected. On external characters it is a female.

Distribution (fig. 19). — Suriname, where known only from creeks in the
northern lowlands, but in this area widely distributed (see list of material),
and Brazil, where known from a single specimen collected at Cudajas, over a
century ago.

Discussion. — The colour description of Centromochlus aulopygius from
Cudajas (= Codajas) by Eigenmann & Eigenmann (1890: 270): “Light
brownish, mottled with darker; a dark lateral band; dorsal fin dark brown;
caudal mottled with darker, a dark longitudinal line on the upper and lower
lobe”, does certainly not suggest Tatia intermedia , the species called C. aulo¬
pygius by Eigenmann (1912), nor does it resemble any other known species
of ,the Amazon basin. The only species which this description fits is Tatia
creutzbergi In addition the statements: “the backward-projecting process
of the dorsal plate simple, not turned outward or inward behind”, and “caudal
emarginate”, are also applicable to T. creutzbergi , and not to any of the other
known species of the genus.

Originally I was informed that the specimen had disappeared, but for¬
tunately it was subsequently found and forwarded. I found my suspicions
of its identity fully confirmed: in the eighty years that have passed since
the Eigenmanns studied the specimen, it has certainly become somewhat
decolourized, but even now there is a broad longitudinal band on each side,
brownish in colour, and also some brown pigment along the upper and lower
edges of the caudal fin and on the dorsal spine. The broad nuchal shield,
and the short anal fin (with only eight rays), as well as the fairly large
adipose fin, reaching to beyond the anal fin, are all characteristic of T. creutz¬
bergi. The caudal fin, however, now gives an impression of being forked, but
that is evidently because the membrane is damaged, for the Eigenmanns
described it as emarginate. All in all, I have no hesitation in referring this
specimen to T . creutzbergi , although fresh material from the region where
it was obtained is desired. Thiis specimen constitutes a remarkable extension
of range of the species, which was hitherto known from streams in the
northern part of Suriname only.

The variation in numbers of anal fin rays is remarkable; in the females it
is normally 8, but in several males I definitely counted 11, the posterior ones
very thin, but perfectly discernible. The sexes agree in having eight sup¬
porting bones (actinosts).

Tatia schultzi (Rossel) (figs, iof, 20)

Centromochlus schultzi Rossel, 1962, Senckenb. Biol., 43: 27 — oberer Rio Xingu,
Brasilien.

Centromochlus perugiae ; P. de Miranda Ribeiro, 1962, Publ. Avuls. Mus. Nac. Rio

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

8l

6

82

ZOOLOGISCHE VERHANDELINGEN l $2 (1974)

de Janeiro, 45: 9 (Rio Araguaya, Estado de Goyaz; Alto rio Xingu); P. de Miranda
Ribeiro, 1964, Bol. Mus. Nac. Rio de Janeiro, Zool., 246: 4 (Rio Araguaya; Alto rio
Xingu).

Material. — One specimen ($), i960, Rio Xingu (H. Schultz, SMF no. 5462),
standard length 60mm, holotype of the species. Four specimens, i960, Rio Araguaya,
Estado de Goiaz (H. Schultz, MRJ no. 9416), standard length 27, 28, 33, 46 mm. Two
specimens, i960, Alto Rio Xingu (H. Schultz, MRJ no. 9417), standard length 33,
6 2]/ 2 mm. One specimen, 30 April 1968, Xaventina, Rio das Mortes (R. H. McConnell,
BM no. 1971.7.29: 3), standard length 28 mm.

Characters. — Size large (standard length 28-6034 mm); nuchal plate
with thin, almost stalked horns, which are curved outwards; number of teeth
on both edges of the pectoral spine similar to that found in the other species,
but especially towards the distal end of the spine these teeth are smaller than
in the majority of species, and along the anterior edge all teeth are antrorse,
whereas in most other species these teeth are retrorse; hind-border of ad-
pressed anal fin about opposite hind-border of adipose fin; number of rays in
the anal fin 9 or 10; colour and colour pattern diagnostic.

Colours. Upper half of head and body lightly pigmented, light brown, the
upper part of the body, and the caudal region also lower down, with not very
numerous round to oval, mostly rather large dark brown blotches.

Distribution. — The upper course of the Rio Xingu, the Rio Araguaia
and its tributary the Rio das Mortes.

Discussion. — As Rossel (1962), who had eight specimens, made no men¬
tion of variation in shape of the anal fin, it appears that all his specimens
were females. The specimens examined by me have also unmodified anal fins.
Unless this species differs from all its congeners by having no sexual dimor¬
phism in shape of the anal fin, the adult male remains to be described.

Rossel has drawn attention to a number of morphological differences
between this species and the superficially somewhat similar T, perugiae. The
most conspicuous differences are in the shape of nuchal shield and horns, and
pectoral spine. Even in colour and pattern there is a clear difference: 7 \
perugiae has the dark blotches closer together; they are pronounced dorsally,
but become vague on the flanks. In T . schultzi on the other hand, the less
numerous dark blotches are all equally dark, those on the lower part of the
caudal peduncle being of the same intensity as those of the back.

Inasmuch as part of the material listed by P. de Miranda Ribeiro (1962b,
1964a) as Centromochlus perugiae is from the type-locality of T. schultzi,
and was obtained by the same collector, I considered it a safe assumption that
it belonged to this species, an assumption that proved right when subsequently
I had an opportunity to examine the specimens in question.

84

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Tatia brunnea species nova (figs, iod, 21)

Material. — Four specimens (4 $ ), 28 December 1963, Kwambaolo Kreek, near Dam
(Boeseman, RMNH no. 26198), standard length 36, 36^4, 40, 42^4 mm. One specimen
(3), 15 March 1966, Compagnie Kreek (Mees, RMNH no. 26196), standard length
57 mm, holotype of the species. Three specimens (r 2 9 ), same data (Mees, RMNH
no. 26197), standard length 47^2 ($), 55, 60 mm. One specimen (9 ?), 14 October 1966,
Sara Kreek, 27 km S. of Dam (Nijssen, ZMA no. 105860), standard length 39 mm.
Four specimens (sex?), 20 October 1966, Gran Kreek, 63 km S. of Affobakka (Nijssen,
ZMA no. 105526), standard length 25, 28, 28, 35 mm. Seven specimens (5 $, 2 ?),
21 April 1967, Maka Kreek, tributary of Lawa Rivier 10 km S. of Stoelman Eiland
(Nijssen, ZMA no. 105849), standard length 26^4, 33, 39, 41, 42, 48, 51 mm. Five speci¬
mens (5 $), 22 April 1967, Soea Kissi Kreek, tributary of the Tapanahoni Rivier, 12 km
S. of Stoelman Eiland (Nijssen, ZMA no. 105847), standard length 41, 51, 52, 54, 55 mm.
16 specimens (4 5,7 $,5 ?), 24 April 1967, Kamaloe or Saloea Kreek, eastern tributary
of the Marowijne, French Guiana (Nijssen, ZMA no. 105851), standard length 28-57 mm.

Characters. — Size large (standard length 25-60 mm); nuchal plate with
horns which are strongly curved outwards, usually club-shaped; hind-border
of anal fin in advance of hind-border of adipose fin; usual number of rays
in the anal fin 8 or 9; male anal fin pointed, with its third and fourth rays
longest; caudal fin more deeply forked than in any other species, when indivi¬
duals of similar size are compared; colour and colour pattern diagnostic.

The male anal fin develops at a standard length of 35 to 40 mm.

Colour. Fresh specimens are dark earth brown in colour, with only vague
paler areas; pectorals, dorsal and adipose fins spotted with brown, the dorsal
fin largely dark brown; ventrals and anal fin hyaline, caudal fin hyaline with
large irregular blackish brown dots. Under surface from chin to ventrals
unpigmented, white. The dorsal shield is usually distinctly paler than die body.
In older specimens some of the colour has gone, and instead of more or less
even brown, the body shows wavy longitudinal bands and dots of dark brown,
alternating with pale brown areas.

Distribution. — Known from the Suriname Rivier and Marowijne basins,
in Suriname and adjacent French Guiana (Kamaloe or Saloea Kreek, see list
of material).

Tatia concolor species nova (figs, iog, 22)

Material. One specimen ($), 17 May 1967, headwaters of Coppename Rivier (3°49'N,
56°57'W), Suriname (Nijssen, ZMA no. 106210), standard length 33 mm, holotype of
the species. Two specimens (3, $), same data (Nijssen, ZMA no. 106209), standard
length 29, 30 mm.

Characters. — Size small (standard length 29-33 mm )» dorsal shield with
horns which are more or less diamond-shaped at the tip; anal fin with 8 rays;
hind-border of male anal fin rounded, reaching to clearly behind hind-border
of adipose fin; colours as described below.

UEES, AUCHENIPTERIDAE AND PIMELODIDAE

85

86

ZOOLOGISCHE VERHANDELINGEN Itf (l 974 )

The two males, at standard lengths of 30 and 33 mm, already have a fully
developed male anal fin, and this has led me to the conclusion that it is, indeed,
a small species, and that the small size of the available material is not due to
immaturity.

Colours. The plainest of all species, dark grey above, white below; some
pigment spots on the chin, but none farther back on the under surface of the
head. The melanophores are very large and everywhere except on the back
widely spaced, giving, even to the naked eye, a mottled appearance to the fish.
Fins hyaline, except for the dark base of D. On D, P, A and C a few scattered
melanophores.

Discussion. The point in the middle of the caudal fork shown by the type
specimen is an individual aberration, as the other two specimens, one of
which is of the same sex as the type, do not have it. The figure does not
bring out sufficiently the fact that the anal fin reaches farther backwards
than the adipose fin.

Tatia galaxias species nova (figs. 13, 23)

Centromochlus auloPygius; Pellegrin, 1899, Bull Mus. Hist. Nat. Paris, 5: 188
(l’Apure, Venezuela); Schultz, 1944, Proc. U.S. Nat. Mus., 94: 240 (Venezuela).

Tatia aulopygia ; (pt.) Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 10
(Apure) ; Mago-Leccia, 1967, Soc. Venez. Cienc. Nat. Bol., 27: 225 (Venezuela).

Material. — One specimen, 1893/1895, TApure, Venezuela (G. Geay, MNP no. 98-34),
standard length 31 mm. One specimen ($), May 1925, Cano de Quiribana into Rio
Orinoco, Venezuela (C. Ternetz, CAS no. 6567), standard length 61 mm, holotype of
the species. Six specimens (all #), same data (C. Ternetz, CAS no. 6568), standard
length 48^2-55^4 mm (two specimens of this sample now RMNH no. 26493).

Characters. — Size large (standard length 31-61 mm); nuchal plate with
horns which are almost as strongly curved outwards as, and approach in
shape those of T . brunnea ; hind-border of adipose fin slightly more posterior
in position than hind-border of adpressed anal fin; usual number of rays
in the anal fin 10 (iii.7); male anal fin in the available material only moder¬
ately modified; caudal fin shallowly forked; colour and colour-pattern diag¬
nostic.

The male anal fin presumably develops at a standard length of about
45 mm since one of the males, of 49 mm standard length, does not yet have
it completely modified.

Colour. Whole body, including the unpaired fins, dark brown with evenly
spaced small white dots; chin with scattered pigment, underparts from chin
to anus, and ventrals unpigmented, pectorals pigmented lightly.

The old MNP-specimen does not show this pattern; its pigment is brown,
with only on the caudal fin remains of some ill-defined pale spots.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

87

88

ZOOLOGISCHE VERHANDELINGEN 13 2 (1974)

Distribution (fig. 13). — Orinoco basin, Venezuela, where known from
the type-locality, Cano de Quiribana, and from the Rio Apure — without exact
locality, but quite likely very near Cano de Quiribana.

Discussion. — The colour-pattern of this species is reminiscent of that
of T. intermedia , but the body is darker, and the spots are much smaller;
also the tail is dark with white spots, whereas the tail of T. intermedia can
better be described as pale with dark spots and cross-bars. Moreover the
horns of the nuchal plate appear to be slightly thinner and more strongly
curved than those of T. intermedia , and I am therefore of the opinion that
the two forms are definitely different.

The MNP-specimen is too decolourized for a certain identification, but
even before the fresh material was received, I had decided that it could not
be identified with any known species, whereas in such morphological charac¬
ters as are available it appears to agree with the fresh material. In addition
its locality Apure is in the same river system as the locality Cano de Quiribana.

T. galaxias is the only species of its genus at present known from Vene¬
zuela.

Tatia punctata species nova (figs. 10c, 24)

Material. — Two specimens ($, 9 ), 20 April 1962, Igarape Kumadueni Fittkau, Rio
Paru de Oeste (= Erepecuru), Brazil near border of Suriname (J. Gery, RMNH
no. 26494), standard length 34 (9), 41 mm. One specimen ( $ ), 17/18 February 1964,
creeks between Kabel and Lombe (Boeseman, RMNH no. 26495), standard length 45 mm,
holotype of the species. Three specimens ($, 2 9 ), same data (Boeseman, RMNH
no. 26496), standard length 36, 39, 39 mm. One specimen (S), 27 November 1965, Tapa-
nahoni Rivier about 2 km upstream from its confluence with the Paloemeu (Mees,
RMNH no. 26497), standard length 34 mm. Three specimens (2 #, 9 ), 20 October 1966,
Gran Kreek ca. 63 km S. of Affobakka (Nijssen, ZMA no. 105525), standard length
35 (2), 38, 42 mm.

Characters. — Of medium size (standard length 34-45 mm); nuchal plate
with blunt horns, moderately bent outwards, about the same as in T. inter¬
media; hind-border of adpressed anal fin behind hind-border of adipose fin,
or almost opposite it; usual number of rays in the anal fin 10 (iii.7); male
anal fin rather thinner than in the other species from Suriname, its tip
rounded, with the third ray longest and strongest; colour and colour-pattern
diagnostic.

A specimen with a standard length of 34 mm already has a well-developed
male anal fin.

Colour. Densely mottled and dotted with dark grey on a pale (unpigmented)
background: the upper surface of the head, from snout to dorsal fin is almost
uniform dark grey; underparts from chin to anus white, unpigmented. In

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

89

90

ZOOLOGISCHE VERHANDELINGEN I$2 (1974)

preserved material the spots tend to become brown, and will probably dis¬
appear entirely in another ten years or so, which would make old material
very difficult to identify.

Distribution. — Suriname: Suriname Rivier basin; Tapanahoni; Brazil:
Upper Paru near Suriname border. It is interesting to know that this species
occurs in the Rio Paru del Oeste, hence in the Amazon basin. Whether it is
widely distributed in Amazonia, or is an originally Guianan species that has
only recently penetrated southwards, remains to be seen.

Tatia reticulata species nova (fig. 25)

Material. — Five specimens, November 1958, Karanambo, Rupununi, British Guiana
(R. H. McConnell), standard length 19, 20, 21, 22, 27 mm, the largest specimen is holotype
of the species (holotype BM no. 1972.7.27: 702, two paratypes BM no. 1972.7.27: 703-704,
and two RMNH no. 26744).

Characters. — Size variation unknown (standard length 19-27 mm, but
probably all specimens are immature); nuchal plate with horns curved out¬
wards as in T. intermedia ; hind-border of adpressed anal fin opposite hind-
border of adipose fin; number of rays in the anal fin 7 or 8; colour and
colour-pattern diagnostic.

All specimens have a normal, unmodified anal fin, but this does not prove
anything about their sex as they are probably immature.

Colours after twelve years of preservation, light pinkish brown, with a
network of white (unpigmented) lines; under surface from chin to anus
white; dorsal spine and basal part of caudal fin lightly pigmented, fins other¬
wise hyaline.

Distribution. — Known from the type-locality only.

Discussion. — Morphologically this species is extremely close to T . inter¬
media and T. punctata , but differs by the lower number of anal rays, and
entirely different colour-pattern.

Tatia simplex species nova (fig. 26)

Material. — One specimen (9 ?), 30 April 1968, Xaventina house beach, Rio das
Mortes, Mato Grosso, Brazil (R. H. McConnell, BM no. 1971.7.29: 5), total length
36^4 mm, standard length 28^ mm, holotype of the species.

Characters. Size variation unknown (only specimen standard length
28 y? mm); dorsal shield with horns curved strongly outwards and ending in
points; pectoral spines with along the outer edge 19, along the inner edge 14
teeth; except for the most distal two or three, which are directed very slightly
outwards, all teeth on the anterior edge are implanted perpendicularly to the
spine, smaller and more widely spaced than is usual in the genus; adipose fin

MEES, AUCHENIFTERIDAE AND PIMELODIDAE

91

On

d

c

§

PQ

3

s

.Q

£

.bp

E

92

ZOOLOGISCHE VERHANDELINGEN I $2 (l 974 )

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

93

extremely small, much smaller than in all other species; hind-border of ad-
pressed anal fin behind hind-border of adipose fin; number of rays in the
anal fin io; colour and colour-pattern as described below.

The available specimen has an entirely unmodified anal fin, but whether
it is a female or a juvenile I cannot say.

Colours. In pigmentation and colour this species is close to T. concolor , but
the dots of pigment are larger, darker (black), and fewer in number.

Distribution. — Known from the type-locality only.

Discussion. — This is a well-differentiated species so that, although only
a single specimen was available, I have felt no hesitation in describing it as
new. Even in pigmentation there are clear differences from T. concolor ,
and in addition there are morphological differences which make it unlikely
that the two are closely related. Whether the exceptionally small adipose fin
is a specific character or an individual aberration must remain uncertain
until more specimens become available.

Genus Glanidium Liitken

Glanidium Liitken, 1874, Oversigt K. Danske Vidensk. Selsk. Forh.: 31 — type by
monotypy, Glanidium albescens Liitken.

Gephyromochlus Hoedeman, 1961, Bull. Aquatic Biol., 2: 135 — subgenus, type by
monotypy Centr onto chlus (Gephyromo chlus) leopardus Hoedeman. Raised without com¬
ment to a full genus by Hoedeman (1968).

Generic diagnosis. — D 1 .5 (I.454), in one species apparently 1 . 6 , A 10-13,
P I.5 or 1 . 6 , V 6, C i. 15.1 plus rudiments. This genus is very close to Tatia ,
but its members attain a distinctly larger size; the anterior borders of the
spines in D and P are either smooth, or provided with small teeth only; the
upper surface of the head is covered with thick skin; the horns of the nuchal
plate are very slender, almost stalked; the number of rays in the pectoral and
anal fins tends to be higher than in Tatia. Lateral line complete, but especially
on the anterior part of the body inconspicuous, almost straight. Pectoral pore
as in Tatia. Sexual dimorphism of the anal fin as in Tatia. The differences
between Glanidium and Tatia are slight and some are not absolute (for
example T. schultzi has the teeth along the anterior edge of the pectoral spine
poorly developed, whereas G. albescens and G. ribeiroi have them fairly well-
developed); nevertheless, I regard it as justified to maintain Glanidium as
a separate genus, in a group so compact and homogeneous.

Distribution (fig. 28). — The genus is widely distributed in eastern South
America, with its point of gravity in south-eastern Brazil.

Discussion. — There have been few previous attempts to define Glanidium .
The first of these was made by Eigenmann & Eigenmann (1890: 266), who

94

ZOOLOGISCHE VERHANDELINGEN I$2 (1974)

Fig. 27. Right hand pectoral spines of Glanidium- species, viewed from above: (a) G.
albescens (UZMK no. 335, s.l. 100 mm), (b) G. ribeiroi (FM no. 54234, paratype, s.l.
101 mm), (c) G. melanopterum (RMNH no. 17294, s.l. 103 mm), (d) G. leopardus (ZMA
no. 105854, s.l. 109 mm), (e) G. catharinensis (MRJ no. 5169, s.l. 88 mm). 4 X.

recognized it as a subgenus only, as follows: “head granular above, Centro-
mochlus; head covered with skin, Glanidium”. Haseman (1911) used Glani-
dium without comment. A. de Miranda Ribeiro (1911) gave a generic diag¬
nosis, but mentioned no characters which would differentiate Glanidium from
T atia.

Seven species have been described, but there are indications that one of
them, G. cesarpintoi, is a synonym of G. ribeiroi , or alternately, that it may
be referable to the genus Tatia, which leaves the genus with six species, one
of which (G. catharinensis) is doubtful. One species occurs in Suriname.

Glanidium albescens Liitken (figs. 27a, 28)

Glanidium albescens Liitken, 1874, Oversigt K. Danske Vidensk. Selsk. Forh.: 31 —
flumina Rio des Velhas cum affluentibus.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

95

Fig. 28. The distribution of the genus Glanidium. 1, G. albescens . 2, G. ribeiroi. 3, G.
melanopterum. 4, G. piresi. 5, G. leopardus. 6, G. catharinensis.

96

ZOOLOGISCHE VERHANDELINGEN 13 2 (1974)

Glanidium albescens ; Liitken, 1875, K. Danske Vidensk. Selsk. Skr., (5) 12: 150,
pi. Ill fig. 5 (Rio das Velhas).

Material. — Six specimens (4 $, 2 $), 1850-1856, Rio das Velhas (J. Reinhardt,
UZMK nos. 335, 336, 337, 338, 341, 342), standard length 61 l / 2l 78 ( 9 ), 93 ( 2 ), 95 i 99.
100 mm, syntypes of the species.

Characters. — D. I.5, A 11-12 (iii.8 to iii.9), P I.5 or 1 . 6 , V 6, C i.is.i and
rudiments. Dorsal spine short, 2.5 to 2.6 times in predorsal length, its ante¬
rior edge with some eight to ten small and inconspicuous teeth; pectoral spines
about 1.6 times in predorsal length, with teeth along the whole length of the
anterior and the posterior edges (fig. 27a), see also Liitken’s (1875) plate.

Male anal fin well-modified: the smallest male (of 61^2 mm standard
length) has it only partly modified.

Colour. The specimens which are not entirely decolourized show irre¬
gularly distributed brown dots of various sizes against a light brown back¬
ground. See further Liitken (1875: 151).

Distribution (fig. 28). — At present only known from the type-locality, the
Rio das Velhas, a tributary of the Rio Sao Francisco, Minas Gerais, Brazil. x )

Discussion. — The name Glanidium albescens has been applied widely to
a species inhabiting rivers flowing east and south-east in the states of Rio
de Janeiro and Sao Paulo. I found, however, that all specimens examined
from this area differ from the type-material in the character of the pectoral
spine, as described and figured, and my impression is that this is a specific
difference.

Geographical evidence supports the opinion expressed above, for whereas
on the east coast the species hitherto identified with G. albescens is not known
to occur further north than the Rio Paraiba (= Rio Parahyba), the mouth
of the Rio Sao Francisco, of which the Rio das Velhas is a tributary, lies over
1300 km away to the north.

Subsequent to the description, the only record of presumed G. albescens
from the Sao Francisco basin was by Fisher (1917: 423), who listed under
this name three specimens from Joazeiro. I asked for a loan and to my sur¬
prise the sample (27-29 November 1907, leg. J. D. Haseman, FM no. 57807)
contained not three, but 17 specimens, standard length 32^-47 mm, which
differ from G. albescens in numerous characters. On a later page these speci¬
mens will be described as representing a new species and a new genus.

1) If Pelzeln (1868-1871: VII and map) is right, Natterer’s Rio das Velhas is an
entirely different river, an affluent of the Paranahyba, and ultimately of the Parana.
Although I am not aware that Natterer collected any fishes in his Rio das Velhas, where
he stayed only one day, I mention it as a possible source of confusion.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

97

Glanidium ribeiroi Haseman (figs. 27b, 28)

Glanidium ribeiroi Haseman, 1911, Ann. Carnegie Mus., 7: 381, pi. 78 — Porto Uniao
da Victoria, Parana, Brazil, from the Rio Iguassu.

Glanidium cesarpintoi R. von Ihering, 1928, Bol. Biol. S. Paulo, 12: 46 — Cachoeira
de Emas (Pirassununga), rio Mogy-guassu, Est. de S. Paulo.

Glanidium ribeiroi; R. von Ihering, 1928, Bol. Biol. S. Paulo, 12: 47 (Rio Iguassu);
Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 11 (rio Iguagu); Fowler, 1951,
Arq. Zool. S. Paulo, 6: 467, fig. 495 (Iguagu); P. de Miranda Ribeiro, 1962, Bol. Mus.
Nac. Rio de Janeiro, Zool., 240: 3, 5, fig. 1 (no locality).

Glanidium albescens; de Aramburu, Aramburu & Ringuelet, 1962, Physis, 23: 229,
fig. 1 (Arroyo Los Dos Hermanos, Iguazu, Misiones); Ringuelet, Aramburu & de
Aramburu, 1967, Los Peces Argentinos de Agua Dulce: 275, fig. 15 B (Los dos Her¬
manos, Iguazu, Misiones).

Material. — Two specimens (both $), 1909, Porto Uniao da Victoria, Rio Iguagu
(J. D. Haseman, FM no. 54243), standard length 101, 102 mm, paratypes of the species.
One specimen ($ or imm.), undated, Arroyo Los Dos Hermanos, Iguazu, Misiones
(collector unknown, MLP no. 1-II-61-6), standard length 67 mm.

Characters. — D I.5, A 10, P 1 .5 or 1 . 6 , V 6, C i-15-i and rudiments. Pre¬
dorsal length 3.2 to 3.25 times in standard length, dorsal spine 2.5, 2.8 and 3.0
times in predorsal length, with some ten small knob-like teeth along most
of its anterior edge in two specimens, almost smooth in the third specimen;
pectoral spines short, 1.7 to 2.0 times in predorsal length, with teeth along their
whole anterior and posterior edges, as in G. albescens , but along the anterior
edge the proximal teeth much smaller, the distal teeth larger than in that
species (fig. 27b). Head and body apparently heavier than in G. albescens .

Colour. The two paratypes examined show some irregular dots of different
sizes, as figured by Haseman (1911: pi. 78); see also the colour description
given by that author. The more recently collected specimen from Misiones
has the same irregular pattern of grey spots on a pale brownish-yellow back¬
ground, and there is a concentration of spots on the distal part of the caudal
fin, giving its tips a mottled blackish appearance.

Distribution (fig. 28). — At present only known from the Rio Iguagu
(= Rio Iguassu), and the Rio Mogi Guagu (assuming that I am right in
synonymizing G. cesarpintoi with G. ribeiroi).

Discussion. — As only three specimens, all females, were examined, and
one should allow for some individual variation, it is possible that some charac¬
ters are not as clear-cut as presented in the diagnosis. The broad and heavy
head and body were also observed by Haseman, this character shows clearly
in the distance between the posterior pair of nostrils, which is 11 mm in the
two paratypes of G. ribeiroi , and only 6 mm in a syntype of G. albescens of
comparable size (no. 338, standard length 99 mm).

On geographical grounds I expected the specimens from Argentina recorded
by de Aramburu, Aramburu & Ringuelet (1962) to belong to this species.

7

98

ZOOLOGISCHE VERHANDELINGEN 13 2 (l 974 )

Their figure shows a rather heavy-bodied fish, and therefore supports this
identification. Admittedly in the description one finds: “Espina pectoral
fuerte, con el borde posterior dentado y el anterior denticulado en el tercio
distal ”, but in this species the whole pectoral spine is encased in thick pig¬
mented skin, through which only the tips of the large teeth of its diistal third
protrude. Without removing this skin, only the distal teeth are visible. I have
examined one of the specimens and found my suspicions concerning its iden¬
tity confirmed.

P. de Miranda Ribeiro (1962a) has suggested that G. cesarpintoi does
not differ from G. ribeiroi, and also that both may be synonyms of G. albes¬
cens Auct. (== G. melanopterum). Although, as mentioned in the discussion
of the following species, P. de Miranda Ribeiro has to a certain extent
confused the species, his argument for synonymizing G. cesarpintoi with
G. ribeiroi is reasonably convincing, and is supported by geographical evi¬
dence, as the type-localities of both are in the Parana basin (fig. 28). A
direct comparison between the two, however, remains to be made. The types
of G. cesarpintoi are lost (Britski, in litt.), and attempts by Professor de
Godoy to obtain material for me from the type-locality, where the species
appears to be very rare, have not yet met with success.

My confidence in the synonymy indicated above has been somewhat
shaken by further information received from Professor de Godoy (in litt.),
who describes the colours of G. cesarpintoi in the living fish as blackish
brown, with many small white spots. This colour pattern suggests Tatia
rather than Glanidium , and I wondered if perhaps two species, a Tatia and
a Glanidium, might occur in the Rio Mogi Guassu, of which the former could
be related to, or identical with, T. neivai. This, however, appears to be not
so, and Professor de Godoy has summarized the differences between G. cesar¬
pintoi and T. neivai as follows:

G. cesarpintoi T. neivai

maxillary barbel 1.5 X the length maxillary barbel 2 X the length of
of the pectoral spine the pectoral spine

length of pectoral spine greater than length of pectoral spine equal to

the distance from its basis to the the distance from its basis to the tip

tip of the snout of the snout

lobes of caudal fin pointed lobes of caudal fin rounded

P, V and A light, almost hyaline P indistinctly maculated black; V

and A with plumbeous base

body with more and smaller white body with fewer and larger white

dots dots

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

99

Until adequate material becomes available, it appears pointless to speculate
further about the identity and affinities of G. cesarpintoi.

Glanidium melanopterum A. de Miranda Ribeiro (figs. 27c, 28)

( Glanidium albescens Auct., nec Liitken)

Glanidium melanopterum A. de Miranda Ribeiro, 1918, Rev. Mus. Paul., 10: 643 —
Piquete, S. Paulo.

Centromochlus (Glanidium) albescens ; Steindachner, 1876, Sitzb. Akad. Wiss. Wien,
Mathem.-Naturw. G., 74 (1) : 664 (Rio Parahyba bei Campos).

Centromochlus albescens ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) 1 : 157 (Rio Parahyba, Rio Janeiro, Macacos); Eigenmann & Eigenmann, 1890,
Occ. Pap. Calif. Acad. Sci., 1: 270 (coast streams from Rio Janeiro to the Amazon);
Chranilov, 1929, Zool. Jahrb. Anat., 51: 332, 376, 377 fig. 11, 427 fig. 23a (no locality,
presumed to refer to this species); van der Stigchel, 1946, South American Nematognathi:
98 (Port Real, Rio de Janeiro); van der Stigchel, 1947, Zool. Meded., 27: 98 (Port
Real, Rio de Janeiro).

Glanidium albescens ; Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 34
(coast streams from Rio Janeiro to the Amazon); A. de Miranda Ribeiro, 1907, A
Lavoura, 11: 186 (Iporanga); (pt.) A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio
de Janeiro, 16: 358 (Rio Parahyba, Macacos (Rio de Janeiro), Iporanga (S. Paulo));
Haseman, 1911, Ann. Carnegie Mus., 7: 382 (Rio Paranahyba (Entre Rios); Porto
Alegre, Rio Grande do Sul; Rio Ribeiro da Iguape); (pt.) Fisher, 1917, Ann. Carnegie
Mus., 11: 423 (Rio Ribeira da Iguape; Serraria Minas, Rio Parahybuna, into Parahyba;
Entre Rios, Rio Parahyba; Porto Alegre); R. von Ihering, 1928, Bol. Biol. S. Paulo,
12: 47 (da vertente atlantica, do Rio de Janeiro ao Amazonas) ; Fowler, 1951, Arq. Zool.
S. Paulo, 6: 466 (distribution compiled, no original information) ; P. de Miranda Ribeiro,
1962, Bol. Mus. Nac. Rio de Janeiro, Zool., 240: 3 (Lagoa Feio, Campos, Ilha das Pombas,
Rio Paraiba no Estado do Rio de Janeiro, Rio Pomba, Cataguazes, no Estado de Minas
Gerais e Lagoa dos Quadros no Estado do Rio Grande do Sul); P. de Miranda Ribeiro,
1962, Publ. Avuls. Mus. Nac. Rio de Janeiro, 45: 9 (Lagoa Feia, Campos, Estado do
Rio de Janeiro; Ilha das Pombas, Rio Parahyba, Estado do Rio de Janeiro; Rio de
Janeiro (Sepitiba?); Rio Parahyba, Estado do Rio de Janeiro; Humboldt, Estado de
Santa Catharina; Rio Pomba, Cachoeira Vista-Alegre, Municipio de Cataguazes, Estado
de Minas-Gerais; Lagoa dos Quadros, Estado do Rio Grande do Sul).

Glanidium ribeiroi; A. de Miranda Ribeiro, 1918, Rev. Mus. Paul., 10: 734 (Iporanga,
Rio Ribeiro) ; P. de Miranda Ribeiro, 1962, Publ. Avuls. Mus. Rio de Janeiro, 45: 10
(Rio Itapocu, Municipio Guara-Mirim, Estado de Santa Catharina).

Glanidium melanopterum ; R. von Ihering, 1928, Bol. Biol. S. Paulo, 12: 47 (Piquete,
vertente do rio Parahyba); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 11
(Sao Paulo); Fowler, 3951, Arq. Zool. S. Paulo 6: 467 (Sao Paulo); P. de Miranda
Ribeiro, 1953, Arq. Mus. Nac. Rio de Janeiro, 42: 391 (no locality); P. de Miranda Ri¬
beiro, 1962, Bol. Mus. Nac. Rio de Janeiro, Zool., 240: 3 (no locality); Britski, 1969,
Pap. Avuls. Zool., 22: 204 (Piquete, Cachoeira, Estado de Sao Paulo).

Glanidium cezarpintoi ; P. de Miranda Ribeiro, 1962, Publ. Avuls. Mus. Rio de Janeiro,
45: 10 (Rio de Janeiro).

Material. — Seven specimens (5 $, 2 9 ), 1864, Macacos (Wm. M. Roberts, MCZ
no. 7626), standard length 91-120 mm. One specimen ( 9 ), 1865, Rio Parahyba (Thayer
Exp., MCZ no. 7366), standard length 120 mm. One specimen ( 9 ), same data (Thayer
Exp., MCZ no. 7367), standard length 93 mm. One specimen (9 ?), same data (Thayer
Exp., MCZ no. 8192), standard length 60 mm. One specimen, 1872, Rio de Janeiro
(Hassler Exp., MCZ no. 7348), standard length 77 mm. Seven specimens, 1890, Port

IOO

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Real, Rio de Janeiro (Hardy dti Dreneuf, RMNH no. 17294), standard length 56^-
103 mm. One specimen ($ ?), January 1902, Piquete, Estado de Sao Paulo (H. Silva,
MZSP no. 345), standard length 128 mm, lectotype of the species (cf. Britski, 1969).
One specimen ($), 22 May 1908, Serraria Minas, Rio Parahybuna, an affluent of Rio
Parahyba (J. D. Haseman, FM no. 57808), standard length 88 mm. Two specimens,
2 June 1908, Entre Rios, Rio Parahyba (J. D. Haseman, FM no. 57809), standard length
105, in mm. One specimen, 19 January 1909, Porto Alegre (J. D. Haseman, FM
no. 57810), standard length 137 mm. 19 specimens, no date, Rio Parahyba (no collector,
MCZ no. 7378), standard length 52-128 mm. Three specimens (2 S, 9) without data
(MCZ no. 36187), standard length 140, 145, 151 (9) mm. One specimen, “Brazilie”,
without further data (RMNH no. 17246), standard length 105 mm.

Characters. — D I.5, A 1011 (usually iii.7 or iii.8), P 1.5 or 1 . 6 , V 1.5,
C i.is.i and rudiments. Dorsal spine 1.8 to 2.4 times in predorsal length, its
anterior edge smooth, or distally with three or four rudimentary teeth; pec¬
toral spines long, 1.25 to 1.4 times in predorsal length, more distinctly curved
backwards than in the other species, with on their anterior edge teeth on the
distal half only (fig. 27c).

Male anal fin perhaps only moderately modified, and only its anterior rays;
its posterior border lunate.

Colour. All specimens were collected long ago, and had been in preservative
for at least sixty years when I examined them. Several had become entirely
decolourized, but those which still showed some colour and pattern were
light earth-brown above, mottled with irregular larger and smaller dark brown
dots. The dorsal fin was well-pigmented, and the posterior half of the caudal
fin appears to have been entirely black, at least in some specimens; the other
fins are largely hyaline, with at most some small dark blotches and scattered
melanophores. Ventral surface, from chin to anal fin, white, unpigmented.

Distribution (fig. 28). — Occurs in rivers of south-eastern Brazil, running
east, from the Rio Parahyba and its tributaries in the north, to Porto Alegre
and Lagoa dos Quadros in the south; this range includes the extreme south
of the state of Minas Gerais, and (the eastern parts of) the states of Rio de
Janeiro, Sao Paulo, Santa Catarina and Rio Grande do Sul. I know of no
records yet from the state of Parana, but the species is certain to occur there
as it has already been found in the Rio Ribeira, which originates in Parana.

Fowler (1951: 466-467) records the species from the Rio Mucuri, Rio
Doce and Rio Jequitinhona, quoting Steindachner (1876b: 664) as his refer¬
ence. On the place indicated, Steindachner wrote, however: “Diese zuerst von
Dr. Liitken nach Exemplaren aus dem Rio das Velhas beschriebene Art kommt
im Rio Parahyba bei Campos sehr haufig vor; aus dem Rio Mucury, Rio
doce, Rio Jequitinhonha, etc. ist sie derzeit noch nicht bekannt, durfte aber
hochst wahrscheinlich daselbst nicht fehlen”. These three rivers must there¬
fore be excluded from the known range of the species.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

IOI

The inclusion of the Amazon in the range of this species, first by Eigen-
mann & Eigenmann (1890), followed by nearly all later authors who gave
its range, was based on the specimens from Macacos. The Eigenmanns (1890:
490, and map) identified Macacos with the stream connecting the Rio Para
with the Amazon, in the delta of the Amazon. A. de Miranda Ribeiro (1911),
however, who was in a better position to know the geography of south¬
eastern Brazil than workers living in other continents, expressly listed Maca¬
cos as being in the state of Rio de Janeiro. For this reason I feel justified in
eliminating the Amazon from the range of the species *).

In the discussion of G. albescens I have demonstrated that the locality
Joazeiro must also be rejected for this species, being based on a different
species. This leaves the species with a limited range in south-eastern Brazil,
defined above.

Discussion. — As indicated in the discusson of G . albescens , this is the
species which in literature has been widely misidentified as G. albescens , a
species known from its type-material only.

The lectotype of G . melanopterum, examined by me, is now entirely
decolourized, except for some dark central rays in the caudal fin: just as in
specimens of various other samples. The pectoral spines are rather short,
1.5 times in predorsal length, but both spines have their extreme tips
broken off.

P. de Miranda Ribeiro (1962a) himself indicated that the specimen from
Rio de Janeiro which he had catalogued as G . cesarpintoi (cf. P. de Miranda
Ribeiro, 1962b), is referable to the present species, and as he expressed doubt
about the validity of G. ribeiroi, it is apparent that he held the same opinion on
the specimen catalogued under that name. On geographical evidence both
specimens must be G. melanopterum .

Glanidium piresi A. de Miranda Ribeiro (fig. 28)

Glanidium piresi A. de Miranda Ribeiro, 1920, Comm. Linh. Telegr. Estr. Matto-Grosso
ao Amazonas, 58, Annexo 5:14 — Rio S. Manoel.

1) The correction of the locality Macacos from the Amazon delta to Rio de Janeiro
affects the distribution of a number of other species of which the range had been extended
to the Amazon delta on the basis of this locality only. Eigenmann & Eigenmann (1890)
listed the following species from Macacos: Rhamdia quelen, Rhamdella minuta, Pimelo -
della buckleyi (?), Centromochlus albescens , Pygidium maculatum (?), Loricaria lima ,
Plecostomus commersonii, Plecostomus vermicularis, Callichthys callichthys.

Two of these species, marked with a query, were probably misidentified; all other
species are well-known to occur in south-eastern Brazil, but with the exception of
Rhamdia quelen , Rhamdella minuta, and Callichthys callichthys , which have a very wide
distribution, can no longer be accepted as occurring in Amazonia. The fact that all
Macacos species are typical of the fauna of south-eastern Brazil, further supports A. de
Miranda Ribeiro’s statement that this locality is in the state of Rio de Janeiro.

102

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Glanidium piresi; R. von Ihering, 1928, Bol. Biol. S. Paulo, 12: 47 (Rio S. Manuel);
Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 11 (Mato Grosso); Fowler,
1951, Arq. Zool. S. Paulo, 6: 467 (Mato Grosso); P. de Miranda Ribeiro, 1953, Arq.
Mus. Nac. Rio de Janeiro, 42: 391 (name only); P. de Miranda Ribeiro, 1962, Bol. Mus.
Nac. Rio de Janeiro, Zool., 240: 3 (name only).

Material. — None.

Characters. — D 1 . 6 , A io. In the first place characterized by its enormous
size (the type was reported to have a length, presumably total length, of
49 cm), also by pectoral spines with teeth on both anterior and posterior
borders, six dorsal rays, and plain coloration *).

Discussion. — This species is known to me only from the original descrip¬
tion, based on a single dry specimen of 49 cm length. No additional material
has been recorded. One would expect the type to be in the museum of Rio
de Janeiro (A. de Miranda Ribeiro, 1920: 3), but neither in his list of types,
nor in the catalogue did P. de Miranda Ribeiro (1953, 1962b) mention it and
Britski (in litt., 3.VIII.1970) has vainly searched for it in the Museu Nacional
do Rio de Janeiro. Anyway, enough characters are given in description and
figures to show that this species is not particularly close to G. leopardus, the
geographically nearest member of its genus.

Glanidium leopardus (Hoedeman) (figs. 27d, 28, pi. 3)

Centromochlus (Gephyromochlus) leopardus Hoedeman, 1961, Bull. Aquatic Biol., 2:
135 — French Guiana: mainland, Litany river, village Aloike.

Gephyromochlus leopardus ; Hoedeman, 1968, Elseviers Aquariumvissen Encycl., 1:
147 (Frans-Guyana).

Material. — One specimen (5), 27 November 1957, Aloike, Litani Rivier (J. Gery,
ZMA no. 102233), standard length 98 mm, holotype of the species. Twelve specimens,
24 April 1967, Makakreek, left shore Lawa Rivier, 10 km south of Stoelmanseiland
(Nijssen, ZMA no. 105854), standard length 53-109 mm.

Characters. — D I.4, A 12-14 (of which 9 or 10 branched), P 1 .5, V 6,
C 15-16 (divided rays). Dorsal spine 2.0 to 2.4 times in predorsal length, its
anterior edge smooth; pectoral spines 1.4 to 1.65 in predorsal length, with
their tip pungent, entirely smooth along their anterior edge, 12-18 (mostly
15-18) teeth on the posterior edge (fig. 27d); tail shallowly forked, its lobes
rounded; male anal fin strongly modified; colour and colour pattern diag¬
nostic.

1) In the drawings of dorsal and ventral aspects of the type, the anterior borders of
the pectoral spines are shown smooth, without teeth, but they are shown in the drawing
of the lateral aspect, and are definitely mentioned in the description.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

103

Colour. Very large and often confluent dark brown blotches on a brownish-
white to white background.

Distribution (fig. 28). — The material listed above represents all that is
known of the species. Both localities are in the Marowijne basin. Hoedeman
listed his specimen as being from French Guiana, but the village of “Aloike”,
recte Alowike, is on the western shore of the Litani Rivier. As the Lawa is
the boundary river, it is obvious that C. leopardns occurs also in French
Guiana, but at present it is not known from outside Suriname.

Discussion. — Its describer (Hoedeman, 1961) realised that this species is
not quite typical of “Centromochlus” (= Tatia), in which he placed it. He
commented as follows: “Despite its rather elongate appearance, quite unlike
other members of the genus, it agrees in most characters with those species,
and I provisionally want to include leopardus in the genus Centromochlus until
larger samples have become available (including females and juveniles).
Though this species is, however, intermediate in a way between Centro¬
mochlus and Auchenipterus, showing resemblances with members of the
family Trichomycteridae ( = Pygiidae), I found justification for establishing
the new subgeneric name Gephyromochlus”.

As Hoedeman did not explain what his opinion quoted here was based on,
I cannot very well discuss it, except to state that I fail to see any close affinity
of G . leopardus to either Auchenipterus or the Trichomycteridae, and that it
appears to be a very normal representative of Glanidium, a genus not men¬
tioned by Hoedeman, and possibly overlooked by him. Gephyromochlus Hoe¬
deman is a synonym of Glanidium Liitken.

Glanidium leopardus , contrary to statements in its original description, is
not very slender; in body-shape and proportions it does not differ appreciably
from Tatia . The type, and only specimen available to Hoedeman, is in a poor
condition; it is strongly shrivelled and may have been dry, or alternately may
have been kept in far too strong spirits. This gives it an appearance of slen¬
derness, not shared by the recent and well-preserved material (compare Hoe-
deman’s figure with the one given here!).

Glanidium catharinensis P. de Miranda Ribeiro (figs. 27c, 28)

Glanidium catharinensis P. de Miranda Ribeiro, 1962 (22 March), Bol. Mus. Nac. Rio
de Janeiro, Zool., 240: 3, fig. 2 — Rio do Brago do Norte, Sao Ludgero, Municipio de
Tubarao, Estado de Santa Catarina.

Glanidium catharinensis ; P. de Miranda Ribeiro, 1962 (28 Aug.), Publ. Avuls. Mus.
Nac. Rio de Janeiro, 45 : 10 (Rio do Bra$o do Norte, Sao Ludgero, Municipio de Tubarao,
Estado de Santa Catharina).

104

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Material. — One specimen ($), 1940, Rio do Brago do Norte, Municipio de Tubarao,
Estado de Santa Catharina (Padre A. Nihnes, MRJ no. 5169), standard length 88 mm,
holotype of the species.

Characters. — D 1.5, A 10, P 1 . 6 , V 6, C i.is.i and rudiments. Predorsal
length 3.25 times in standard length, dorsal spine 2.6 times in predorsal
length, along its anterior edge five very small teeth, otherwise smooth; pec¬
toral spines 1.75 times in predorsal length, only slightly curved, with teeth
along the whole length of the anterior and posterior edges, but along the
anterior edge the proximal teeth are small (fig. 27c) ; allowing for a certain
amount of individual variation, the pectoral spine agrees with that of G.
ribeiroi ; anal fin ( 6 ) strongly modified; the distance between the posterior
nostrils is 9 mm.

Colour. The preserved specimen is yellowish-brown, with a number of
larger and smaller blackish spots on the body; on the caudal fin these spots
are smaller and rather more numerous.

Distribution (fig. 28). — This species is known from its single type-speci¬
men, collected in the Rio do Bra^o do Norte, Santa Catarina.

Discussion. — Before I had examined the specimen on which G. catha-
rinensis was based, I considered it a possibility that it would belong to G.
melanopterum, as the type-locality is a river flowing east, entirely within the
known range of G. melanopterum . In this opinion, however, I was mistaken;
the specimen does definitely not belong to G. melanopterum , but is very close
to G. ribeiroi and if only it had been collected in the Parana-La Plata basin,
I would not have hesitated to synonymise it with that species. As it is, the
specimen presents a zoogeographical problem, for elsewhere the three closely
related species G. albescens , G. ribeiroi and G. melanopterum are clearly
allopatric. On the other hand, I realise very well that the geographical element
has to be used with caution: from the Suriname Rivier basin four very similar
species of the closely related genus Tatia are known, and therefore the
possibility that in Santa Catarina two very similar species of Glanidium occur
in one river system, cannot be dismissed (G. melanopterum has not actually
been recorded from the Rio do Brago do Norte, but is known from rivers to
the north as well as to the south). Evidently more field-work and collecting
is required before the status of G. catharinensis can be properly evaluated,
and therefore I prefer to recognize it tentatively, although I suspect strongly
that it is a synonym of G. ribeiroi. When P. de Miranda Ribeiro (1962a)
described G. catharinensis , he had no material of G. ribeiroi , for the specimen
which he believed to be referable to it, was actually G. melanopterum , and on
the basis of this misidentified specimen he concluded that G. ribeiroi does not
differ from G. albescens (auct., nec Liitken = G. melanopterum).

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

105

Pseudotatia genus novum

Generic diagnosis. — Close to Tatia and Glanidium , and of the same general
body-shape, but all fins except caudal with more rays; there is a distinct,
slender caudal peduncle; the anal fin is emarginate in outline; males do not
have a modified anal fin; pectoral pore distinct, behind pectoral base.

Type and only known species: Pseudotatia parua n. sp.

Distribution. — Known from the type-locality only: the lower course of
the Rio Sao Francisco, on the border of Pernambuco and Bahia, Brazil.

Discussion. — In the group of genera to which Pseudotatia belongs, differ¬
ences in fin-ray numbers, as well as the presence or absence of a distinct
caudal peduncle, are in my opinion clearly of generic value. To force this
species into the genus Tatia , of which the known species show no differentia¬
tion in fin-ray numbers (except for a slight variation in anal rays), would
disrupt that extremely uniform genus.

Pseudotatia parva species nova (fig. 29)

Glanidium albescens ; (pt.) Fisher, 1917, Ann. Carnegie Mus., 11: 423 (Joazeiro, Rio
Sao Francisco).

Material. — One specimen, 27 to 29 November 1907, Joazeiro, Brazil (J. D. Haseman,
FM no. 70580), standard length 46 mm, holotype of the species. 16 specimens, same data
(FM no. 57803), standard length 32^-46 mm (three specimens of this sample now
RMNH no. 26600).

Diagnostic characters. — D 1 . 6 , A 15-16, P 1 . 6 , V 7, C i.i5-i and rudiments.
Differential characters have been given in the generic diagnosis.

Description. — An apparently small species (standard length 32^2-46 mm)
of normal, subcylindrical body-shape. Head of moderate size, 3.5-3.7 times
in standard length, greatest width of body, between the cleithra, 4.25-4.8
times in standard length, height of head at the same place about 0.8 times its
width, depth of body 4.0-5.2 times in standard length, predorsal length 2.95-3.1
times in standard length, distance from tip of snout to base of ventrals 1.85-2.0
times in standard length.

Snout blunt, rounded, anterior profile of head convex, postoccipital outline
very slightly concave; bones of head and nape covered with thin skin; depres¬
sion containing the fontanel a large oval between the eyes; fontanel elongate,
almost slit-like; horns of nuchal plate flung outwards and downwards,
pointed; each jaw with a single broad band of teeth, no teeth on vomer and
palatines; eyes large, their greatest diameter 2.8-3.6 times in head, 0.5 to 0.75
in snout, and about 1.0 in bony interorbital; posterior pair of nostrils in a
dorsal position, over the anterior halves of the eyes, anterior pair in front
of the middle of the eyes, on the vertical part of the snout; maxillary barbels

io6

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

X

a,

a

,b£

E

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

107

reaching to the end of the head (posterior border of operculum), outer pair
of mandibular barbels about 0.5 to 0.7 times their length, not reaching or only
just reaching to base of pectorals; inner pair of mandibular barbels short,
only about half the length of the outer pair; postcleithral process well-
developed, long and slender; just below this process the pectoral pore.

Dorsal fin with one spine and six rays; the spine of moderate length,
1.7-1.95 times in predorsal length, its posterior border smooth, its anterior
border with some 15-20 small teeth which are reasonably well-developed and
well-spaced on the distal part, but become smaller, irregularly placed and
difficult to count towards the base; a filament of about one-fourth its length
is attached to the spine; the first ray equals in length the spine with filament;
the following rays are successively smaller, and the last one measures from
one-third to two-fifths of the length of the first one.

Anal fin well-developed, with fifteen or sixteen rays, the first two or more
often three rays simple, the others divided; the anterior rays are equal in
length to the longest dorsal ray, or only a little shorter, the posterior rays are
much shorter, and the outline of the fin is more or less concave.

Pectoral fins with one strong spine and six rays; the spine measures
1.25-1.5 times in predorsal length, it is slightly curved backwards, with along
its anterior border some 15-20 small teeth, which become indistinct towards
its base, along the posterior border 10-14 much larger teeth, curved inwards;
the first ray is only a little shorter than the spine, but the succeeding rays
diminish rapidly.

The ventrals have one simple and six divided rays; they are bluntly pointed
rather than rounded in outline when spread. One specimen (out of 17) has
the ventrals with eight rays: the last ray is rudimentary, but perfectly dis¬
cernible, and clearly separate from the penultimate one.

The caudal fin is deeply forked, with acute lobes; it consists of fifteen
divided rays, with above and below one fully developed simple ray, and
some shorter simple rays and rudiments.

The small adipose fin is implanted over the posterior part of the base of
the anal fin.

The lateral line is well developed, almost straight, with a clear arc in its
anterior part.

Sexual dimorphism is almost absent, but in presumed females the uro¬
genital opening is just before the origin of the anal fin, whereas in presumed
males it is at the end of a tube along the anterior edge of the anal fin.

Colour. The material is old and undoubtedly bleached out; it is whitish in
colour, the dorsal surface light brownish, with dispersed large melanophores,
not forming a pattern; all fins hyaline.

108 ZOOLOGISCHE VERHANDELINGEN 1 32 (1974)

Tocantinsia genus novum

Generic diagnosis. — Similar to other genera of Auchenipteridae, but head
broad and depressed, width between the cleithra about twice its depth at the
same place; mouth very wide; whole anterior part of body flattened; pectoral
pore fairly large, over posterior half of pectoral base; six rays in the dorsal
fin, seven rays in the ventral fins, caudal peduncle slender, deeper than wide,
caudal fin deeply forked, with acute tips. Type and only known species:
Tocantinsia depressa n. sp.

Distribution. — Only known from the upper course of the Rio Tocantins.

Discussion. — The species on which this genus is based, is evidently close
to the other genera of Auchenipteridae here treated, but differs in a
remarkable way from all known members of its family as indicated above.

Tocantinsia depressa species nova (fig. 30)

Material. — One specimen, 31 January 1924, Tocantins near Piexa, Goyaz, Brazil (C.
Ternetz, CAS no. 6590), standard length 71 mm. One specimen, 11 February 1924,
Tocantins near Porto National, Goyaz, Brazil (C. Ternetz, CAS no. 6573), standard
length 101 mm, holotype of the species. Two specimens, same sample, standard length
75, 80 mm (one of these now RMNH no. 26498, the other re-registered as CAS
no. 13398).

Characters. — D 1 . 6 , A 12, P 1.6 or I.7, V 7, C i.i5.i and rudiments. Differ¬
ential characters from Tatia and other genera have been given in the generic
diagnosis.

Description. — A presumably fairly small species (standard length of
the four known specimens 71-101 mm), distinguished from Tatia and other
related genera by the broad and flattened anterior part of the body. Pre¬
dorsal length 2.6-2.75 times in standard length; greatest width of body, be¬
tween cleithra, 34-3.6 times in standard length; depth of head at the same
place half its width; distance from tip of snout to base of ventrals 1.5 times
in standard length.

Head large, depressed; greatest width between the cleithra; the depth of
the head at that place is exactly half its width; under surface of the head
flat; its profile on the snout slightly convex, but almost straight from above
the eye to the base of the dorsal fin. Whole upper surface of head and nuchal
region bony, with the exception of an oval depression, containing the fon¬
tanels, between the eyes; horns of nuchal plate reaching to beyond the origin
of the dorsal fin, and flung outwards, club-shaped; the bony part of head
and nuchal region covered by skin. Mouth terminal, the jaws evenly rounded,
the wide gape reaching to below the anterior rim of the eyes; each jaw with
a band of well-developed depressible teeth; no teeth on vomer or palatines;

MEES, AUCHENIPTERIDAE AND PIMELODIDAE 109

Fig. 30. Tocantinsia depressa (CAS no. 6573, s.l. 101 mm).

no

ZOOLOGISCHE VERHANDELINGEN 13 2 (1974)

tongue broad, not free. Eyes small, lateral in position, entirely covered by
skin, their longest diameter 5.5-6.5 times in head (measured from tip of
snout to hind-border of opercle); eyes about 1.5 times in snout (the eyes
being entirely covered by skin, exact measuring is not possible). Nostrils
well separated; the posterior one dorsal in position, above and median from
the anterior part of the eye, large, roundish, with a velum which is well-
developed along the anterior rim, and is interrupted over a short distance
along the posterior rim; anterior nostril straight in front of the posterior one
so that the shortest lines connecting the four nostrils form, viewed from
above, a rectangle; the distance separating anterior and posterior nostril is
half the distance separating anterior from anterior and posterior from
posterior nostril; the anterior nostril tubular, a little behind the upper lip.
Maxillary barbels originating below and just in front of the anterior border
of the eye, and reaching to the pectoral base or a little beyond; outer pair
of mandibular barbels reaching to the end of the mandibular bone (which is
continued much behind the gape), the inner and anterior pair not reaching to
the base of the outer pair. Gill-opening restricted, membranes attached to
the cleithrum, branchiostegals present, opercle moderately developed, with
soft hind-border, as in Tatia and Glanidium.

Lateral line complete, straight or almost straight, continued to the caudal
base.

Dorsal fin with one spine and six rays; the distances between the rays
decrease backwards. The spine is strong and comparatively thick, its length
half the predorsal length of the fish or a little less, with along its anterior
edge a series of about a dozen small antrorse teeth, its posterior border
smooth. A soft filament of about one-fourth of its length is attached to the
tip of the spine; the first soft ray attains about the same length as the spine
with filament, the succeeding rays become progressively shorter, the last one
being about two-fifths the length of the first. On its basal part the fin is
covered with thick skin.

Anal fin more or less square in outline, with twelve rays, the first three
of which are simple, the other nine divided.

Pectorals with one spine and six or seven rays. The spine is long and
strong, with a slight curve backwards; its length 1.5-1.7 times in predorsal
length; along both its anterior and its posterior border, over practically its
whole length, a series of antrorse teeth, numbering 15 to 20 in each series,
usually one or two more along the anterior than along the posterior border.
Compared with the spine, the rays are inconspicuous; the first one is almost
as long as the spine, the succeeding ones very much shorter.

Ventral fins well-developed, reaching to the base of the anal fin, with

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

III

'seven rays, the first one of which is simple, the others divided; the fins
rounded-lanceolate, with the second and third rays longest.

Caudal fin deeply forked with pointed lobes, the central rays about half
the length of the outer rays, with fifteen divided rays, on each side one fully
developed simple ray, and a number of smaller and rudimentary simple rays.

Adipose fin present, small, opposite the anal fin.

All four specimens are similar, and show no differences which could be
interpreted as sexual dimorphism. Although the largest specimen was dissected
and its intestines proved to be in a fair condition, I was unable to find its
gonads, and it remains possible that either all my specimens are of one sex,
or they are immature and for that reason not differentiated.

Colours. Evidently nothing can be said about colours in life; the preserved
specimens are light brown above, and white, unpigmented on the under sur¬
face. The sides of the head and the flanks have a peppered appearance caused
by numerous spots of pigment on a pale background. It is likely that also in
life the coloration is plain, as there remains no indication of any kind of
pattern.

Habitat. — Unknown, but the flat under surface suggests a bottom fish.

Trachelyichthys genus novum

Generic diagnosis. — No adipose fin. In general appearance, with its short
head, compressed body, and long anal fin, close to Trachelyopterus, but differs
by having io-rayed ventrals, a truncate caudal fin, and much larger eyes.
Pectoral pore distinctly behind pectoral base. Type and only known species:
Trachelyichthys decaradiatus n. sp.

Distribution. — At present known from the unique type-specimen only,
collected in the Rupununi River, British Guiana.

Discussion. — The family Auchenipteridae contains but very few species
without an adipose fin. Previously, three or four species were known, divided
over three genera as follows: Trachelyopterus (two species of which one
very doubtful: T. coriaceus Valenciennes, 1840, and T. maculosus Eigen-
mann & Eigenmann, 1888), Trachelyopterichthys (monotypic: T. taeniatus
(Kner, 1858)), and Epapterus (monotypic: E. dispilurus Cope, 1878).
Although from the nomenclatural and systematic point of view I am not in
favour of a proliferation of monotypic genera, the new species presents such
characters that I consider the creation of an own genus for its reception as
the only sound solution. As indicated in the diagnosis, I believe Trachelyopte¬
rus to be its closest relative, but in view of the value I have attached elsewhere
in the Auchenipteridae to the number of ventral rays, it would be inconsistent
to force T. decaradiatus into that genus.

112

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

It is to the point to mention here that, within the Auchenipteridae, the
genera lacking an adipose fin are probably not a natural group. I consider it
likely that Epapterus is closely related to Auchenipterus and Psend epap terns;
the adipose fin in those two genera is so small that it could with some justi¬
fication be called rudimentary.

Trachelyichthys decaradiatus species nova (fig. 31)

Material. — One specimen, 10/11 April 1957, Karanambo area, Rupununi, British
Guiana (R. H. McConnell, BM no. 1971.7.29: 30), standard length 56^2 mm, total length
71 mm, holotype of the species.

Characters. — D I.4, A 35 (ii. 33), P 1 . 6 , V 10 (i.9), C i.is.i and rudi¬
ments. Differential characters have been given in the key to the genera and
in the generic diagnosis.

Description. — A presumably small species with a short head, about as
wide as deep, and for a member of the Auchenipteridae a comparatively
deep, but strongly compressed body. Head 3.9 times in standard length,
greatest width of body, between the cleithra, 3.75 times in standard length,
height of the head at the same place about 0.8 times its width, depth of body
3.4 times in standard length, depth of body above the ventral base three times
its width at the same place, predorsal length 3.45 times in standard length,
distance from tip of snout to origin of ventrals 2.4 times in standard length.

Snout blunt, profile of head ascending to the dorsal spine; bones of head
and nape covered with veiy thin skin, not concealing the underlying rugosity;
depression containing the fontanel a large oval between the eyes, with the
blunter end backwards, the narrower end reaching to the premaxillaries;
fontanel in the middle of the depression, comparatively large, roundish; horns
of nuchal plate rather small, scarcely bent downwards; mouth terminal, gape
about half the body width, upper and lower jaw each with a band of small
teeth of even width throughout, without backward projections, no teeth on
vomer and palatines; eyes large, lateral in position, their greatest diameter
about 3.3 times in head, 0.6 times in snout and 1.7 times in bony interorbital;
posterior pair of nostrils in a dorsal position, over the anterior halves of the
eyes, anterior pair in front of the middle of the eyes, on the vertical part of
the snout; maxillary barbels reaching to a vertical through the dorsal origin
and not quite to the end of the postcleithrum; outer pair of mandibular barbels
reaching almost as far backwards as the maxillary barbels, inner pair of
mandibular barbels about 0.4 times the length of the outer pair; post-
cleithral process well-developed, remarkably broad (see figure); pectoral
pore below it, well behind pectoral base.

8

114 ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Dorsal fin with one spine and four rays; the tip of the spine is missing,
but it appears to have been of moderate length and certainly shorter than
the pectoral spines; its anterior edge is rough, its posterior edge carries small
teeth; the first ray is about two-thirds of the predorsal length, the last one
less than half the length of the first one.

Anal fin long, its base about 2.5 times in standard length, with 35 rays,
the first two of which are simple, the others divided; with the exception of
the short first one, all rays are very much of the same length.

Pectoral fins with one strong spine and six rays, the spine measures 1.6
times in predorsal length, it is very slightly curved, with along its anterior
edge 15 small retrorse teeth, along its posterior border 13 larger antrorse
ones; the first ray is a little longer than the spine, the following rays are
successively shorter.

Ventrals with one simple and nine divided rays, distally rounded, with the
central rays longest, and reaching to just beyond the anal origin; the longest
rays 2.3 times in predorsal length.

The caudal fin has fifteen divided rays, with above and below a fully
developed simple ray and a number of rudiments; it is obliquely truncate
with the upper part slightly the longer.

The lateral line is developed over its whole length, but not conspicuous,
slightly wavy in its course and giving off short branches.

Colour in preservative (14 years after collecting), pinkish brown, darker
on head and back, with an irregular longitudinal series of ill-defined blackish
or brownish blotches on the sides, and some fainter ones scattered elsewhere
on head and body and on the tail.

PlMELODIDAE

The Pimelodidae form by far the largest and most diversified family of
Neotropical catfishes. Gosline (1945) listed no less than 56 genera, and
although not all of these are valid, additional genera have been described
since, and doubtlessly more remain to be discovered; the number of known
species is over 300. It is debatable whether the family as at present under¬
stood is a natural one, and its large number of often very distinctive genera
makes it difficult to define, but Regan (1911: 572) stated: “This family ...
includes a number of diverse types, but is undoubtedly a natural group”, and
I agree. It is also likely, as Gosline (1941: 83) has pointed out, that within
this family the genera without a free eye-rim are not a monophyletic group.
Usually several subfamilies are recognised: Calophysinae, Luciopimelodinae,
Pimelodinae, Sorubiminae. However, as Gomes (1956) pointed out, the cha¬
racters on which one of these subfamilies, the Luciopimelodinae, was based,

MEES, AUCHENIPTERIDAE AND PIMELODIDAE IIS

are also found in a number of genera currently placed in the Pimelodinae.
Gearly the family as a whole has to be known much better, anatomically and
in other respects, before a meaningful division into subfamilies can be
attempted.

Naked catfishes, with the anterior and posterior nares well separated (as
opposed to approximate in the Ariidae, which family includes some super¬
ficially rather similar species); three pairs of barbels: one pair of maxillary
baibels, varying in length from not reaching the base of the pectorals to
reaching well beyond the tip of the tail, and two pairs of mandibular bar¬
bels; occipital process either present and meeting the dorsal plate, or not
meeting the dorsal plate, or absent, never with horns beyond the origin of
the dorsal fin as in the Auchenipteridae; bones forming the upper part of
the skull either clearly visible, covered with thin skin only, or entirely con¬
cealed by thick skin; teeth in bands in both jaws; teeth on vomer and pala¬
tines either present or absent (present in only three of the genera occurring
in Suriname: Hemisorubim, Pseudoplatystoma, and Brachyplatystoma ); eyes
varying from large to small, either with a free rim or, especially in the small¬
eyed forms, without a free rim; eyes absent in some extra-limital forms,
one of which will be discussed in this paper; gill-openings large, membranes
free from the isthmus and from each other, or only narrowly connected;
pectoral pore present in some genera, absent in others: occasionally there
is even variation within one genus; dorsal and pectoral spines present or
absent, if present strong or weak, strongly serrated or almost smooth; adipose
fin present, well-developed and often large; dorsal fin usually with one spine
or simple ray and five to seven divided rays; ventral fins with six rays, of
which the first one is simple, the others divided; anal fin varying from
having no more than eight to over thirty rays; caudal fin of any shape.
Lateral line generally complete, and functional over its whole length, slightly
curved anteriorly or almost straight; in one genus, Microglanis, its posterior
half is more or less rudimentary.

In this family the pectoral pore is of even less systematic significance than
in the Auchenipteridae, for even amongst members of one genus (which
by other morphological characters are closely related), it may be present in
one species, lacking in another (cf. Pimelodus, Pseudopimelodus) . If in the
generic diagnoses no mention is made of a pectoral pore, it means that I have
been unable to find one.

From Suriname, 21 species are known, divided over ten genera as follows:
Hemisorubim (one species), Brachyplatystoma (two species), Pseudo platy-
stoma (one species), Pimelodus (three species), Pimelodella (two species),
Rhamdia (one species), Imparfinis (one species), Heptapterus (five species),

Il6 ZOOLOGISCHE VERHANDELINGEN 1 3 2 (I974)

Pseudopimelodus (three species, one of which occurs in two subspecies) and
Microglanis (two species).

Hoedeman (1961: 135) listed a species of yet another genus from Alowike:
Pinirampus pinirampu (Spix). Reasons for including specimens collected in
the Litani Rivier near Alowike in the fauna of Suriname have been given
under the heading Distribution of Glanidium leopardus. As Pinirampus pini¬
rampu is known from the Amazon basin as well as from British Guiana and
Venezuela, its occurrence in Suriname can be expected; nevertheless I cannot
accept Hoedeman's record, for examination of his specimen revealed that
it is not Pinirampus pinirampu, but Rhamdia quelen. For some additional
particulars see under that species.

Key to the genera of Pimelodidae known from Suriname *)

1 a — Large patches of teeth on vomer and palatines, large species . 2

b — No teeth on vomer and palatines.4

2a — Head strongly depressed; width between the cleithra twice the depth
at the same place; lower jaw strongly protruding; eyes large, less than
seven times in head, less than three times in snout, maxillary barbels
well-developed, but not reaching caudal base; gill-rakers on outer
branchial arch not numerous (8 or 9); predorsal length about 2.1-2.3
times in standard length; caudal fin scalloped or slightly forked, with
rounded lobes; body plain with along the flanks a few conspicuous
black dots, or with round brownish dots of various sizes all over

. Hemisorubim

b — Head moderately depressed; width between the cleithra no more than
one-and-a-half times the depth at the same place; upper jaw pro¬
truding; eyes small, over four times in snout, over ten times in head

. 3

3a — Maxillary barbels long, reaching beyond caudal base (an excellent
character in specimens of up to at least 40 cm standard length, but in
very large specimens the barbels are relatively much shorter); gill-
rakers on outer branchial arch numerous (17-31); predorsal length
2.5*2.8 times in standard length; caudal fin deeply forked, both outer
rays (above and below) elongated into a filament; body plain, without
markings, or with some vague brown spots . Brachyplatystoma

1) This key is made for the identification of species known from Suriname; extra-
limital members of the same genera would not necessarily key out. For example, some
extra-limital species of Imparfinis have maxillary barbels and adipose fin much longer
than here indicated, and at least one species of Pimelodus (P. albicans) has vomerine
teeth.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

II 7

b — Maxillary barbels of moderate length, not reaching beyond anal
origin; gill-rakers on outer branchial arch few (about 4 and a few
rudiments); predorsal length 2.4 times in standard length; caudal
fin scalloped or shallowly forked, with rounded lobes; body with a

striking pattern of black loops and cross-bands Pseudoplatystoma

4a — Orbital rim free.5

b — No free orbital rim.8

5a — Postoccipital process distinctly connected with the predorsal plate

.6

b — Postoccipital process not meeting the predorsal plate . . 7

6a — Postoccipital process more or less triangular, with a broad base and
tapering towards the heart-shaped predorsal plate, which it meets

. Pimelodus

b — Postoccipital process narrow, of equal width throughout Pimelodella
7a — Postoccipital process well-developed, covering about two-thirds of
the distance from occiput to predorsal plate; maxillary barbels long,
reaching at least to below the dorsal fin and usually much farther
backwards; pectoral spines well-developed and pungent; ventrals im¬
planted below or behind posterior border of dorsal base, adipose fin
long, less than 3.5 times in standard length . . . Rhamdia

b —* Postoccipital process rudimentary; maxillary barbels short, reaching
to pectoral base or a little beyond; pectoral spines weak and incon¬
spicuous; ventrals implanted in advance of posterior border of dorsal
base; adipose fin short, over 4.5 times in standard length Imparfinis
8a — Long, slender species, without spines in D and P . Heptapterus
b — Species usually with blunt, heavy, more or less depressed heads, spines

of D and P strong.9

c — In small specimens of Imparfinis the free eye-rim may be indistinct;
the spines are weak and short, continued as soft rays, and may be over¬
looked, so that they might key out with Heptapterus ; check descrip¬
tion of Imparfinis .

9a — Maxillary teeth laterally with a backward projection, species often

large. Pseudopimelodus

b — Maxillary teeth laterally without a backward projection, species small
(maximum standard length in Suriname 38 mm; extra-limital species
to about 70 mm). Microglanis

Genus Hemisorubim Bleeker

Hemisorubim Bleeker, 1862, Atlas Ichth., 2: 10 — type by original designation, Platy-
stoma platyrhynchos Valenciennes.

Il8 ZOOLOGISCHE VERHANDELINGEN 1 32 (1974)

Generic diagnosis. — This monotypic genus is characterized by the
strongly depressed head and body; width between the cleithra twice the depth
at the same place; under surface from snout to anus flat; eyes large, dorsal
in position, with a free rim; snout long, broad, strongly depressed, with a
wide mouth; lower jaw protruding; teeth small, a hand in each jaw, and
patches on the palatine (fig. 32a); postoccipital process well-developed,
tapering, its length about twice its width at the base, and meeting the slender
predorsal plate, which is slightly longer than the postoccipital process; dorsal
and pectoral spines well-developed; maxillary barbels reaching to adipose fin;
pectoral pore above implantation of pectoral rays.

Distribution. — Northern and eastern South America from the Orinoco
to the La Plata basin.

Fig. 32. Shape of tooth-patches in the roof of the mouth of (a) Hemisorubim platy-
rhynchos, (b) Brachyplatystoma vaillantii, (c) Pseudoplatystoma fasciatum.

Hemisorubim platyrhynchos (Valenciennes) (fig. 32a)

Platystoma platyrhynchos Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat.
Poiss. (4 0 ed.), 15: 20 — no locality = Brazil.

Platy stoma platyrhynchos; Castelnau, 1855, Anim Nouv. l’Amer. Sud, Poiss.: 40
(l’Amazone).

Platy stoma platyrhynchus ; Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw.
Cl., 26 (1857) : 398 (Barra do Rio negro); Kappler, 1885, Das Ausland, 58: 919 (Hol-
landisch Guiana); Kappler, 1887, Surinam: 152 (Surinam).

Hemisorubim platyrhynchos ; Bleeker, 1862, Atlas Ichth., 2: 10 (name only); Bleeker,
1863, Ned. Tijdschr. Dierk., 1: 97 (name only); Eigenmann & Eigenmann, 1888, Proc.
Calif. Acad. Sci., (2) 1: 138 (Manacapuru; Rio Negro; Montalegre; Rio Puty; Lago
Alexo; Obidos; Coary; Tabatinga; Hyavary; Tonantins; Sao Paolo); Eigenmann &
Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1: 206 (Orinoco, Amazons, Paranahyba
and their tributaties) ; Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 31
(Orinoco; Amazons; Paranahyba); Eigenmann & Kennedy, 1903, Proc. Acad. Nat
Sci. Philad., 55: 499 (Asuncion, Rio Paraguay) ; Eigenmann, McAtee & Ward, 1907,
Ann. Carnegie Mus., 4: 116 (Paraguay: Corumba); Eigenmann, 1910, Rep. Princeton
Univ. Exp. Patagonia, 3: 391 (Orinoco; Amazons; Paranahyba; Paraguay); Eigen¬
mann, 1912, Mem. Carnegie Mus., 5: 181 (British Guiana) ; Bertoni, 1914, Fauna Para-
guaya: 7 (Alto Parana); Fowler, 1914, Proc. Acad. Nat. Sci. Philad., 66: 263 (Rupu-
nuni River) ; Fowler, 1915, Proc. Acad. Nat. Sci. Philad., 67: 218 (Peruvian Amazon);

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

119

Fisher, 1917, Ann. Carnegie Mus., n: 418 (Santarem; Laguna near Rio Mamore,
Berlin, Bolivia; Lagoa de Paranagua; Caceres); La Monte, 1935, Amer. Mus. Novit.,
784: 6 (Rio Purus and Rio Jurua); Fowler, 1940, Proc. Acad. Nat. Sci. Philad., 91: 224
(Ucayali River basin, Peru); Fowler, 1940, Proc. Acad. Nat. Sci. Philad., 92: 95
(Laguna near Rio Mamore, Berlin); Fowler, 1941, Proc. Acad. Nat. Sci. Philad., 93:
139 (Therezina, Piauhy); Fowler, 1941, Bol. Mus. “Javier Prado”, 5: 380, 381 fig. 19
(Peru: Pebas, Contamana) ; Eigenmann & Allen, 1942, Fish. W. South America: 114
(Iquitos; Manaos; Contamana; Lago Cashiboya); Schultz, 1944, Proc. U.S. Nat. Mus.,
94: 228 (Venezuela); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 51 (Ori¬
noco ; Amazonas; Paraguai; Paranaiba); Fowler, 1945, Los Peces del Peru: 48 (Peru;
Contamana); Bertin & Esteve, 1950, Cat. Types Poiss. Mus. Paris, 5: 58 (no locality);
Fowler, 1951, Arq. Zool. S. Paulo, 6: 588 (Amazonia, Peru, Rio Parnaiba, Sao Paulo,
Paraguay, Venezuela); Boeseman, 1952, Zool. Meded., 31: 181 (Surinam River, near
Joden savanna); Lowe, 1964, J. Linn. Soc. Lond., Zool., 45: no, 141 (Rupununi District
of British Guiana); Ringuelet, Aramburu & de Aramburu, 1967, Los Peces Argentinos
de Agua Dulce: 335 (Argentina: Rosario; Estancia Laguna Oca, Fsa.; Riacho For¬
mosa; Arroyo El Manantial, Tucuman; Laguna Salada, Fsa.; Riacho Ingles, Fsa.).

Hemisorubim platyrhynckus , Gunther, 1864, Cat. Fish. Brit. Mus., 5: 109 (Brazil) ;
Peters, 1877, Mber. Akad. Wiss. Berlin: 470 (Venezuela; Calabozo) ; Cope, 1878, Proc.
Amer. Philos. Soc., 17: 674 (Peruvian Amazon); Vaillant, 1880, Bull. Soc. Philomat.,
(7) 4: 152 (Calderon (Haute-Amazone)); Boulenger, 1896, Trans. Zool. Soc. Lond., 14:
27 (Paraguay); A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 333
(distribution); A. de Miranda Ribeiro, 1918, Rev. Mus. Paul., 10: 733 (Rio Jurua;
Piracicaba); Pearson, 1937, Proc. Calif. Acad. Sci., (4) 23: hi (Mamore and Paraguay
basins); P. de Miranda Ribeiro, 1964, Bol. Mus. Nac. Rio de Janeiro, Zool., 246: 3
(Rio Araguaya); Lundberg & Baskin, 1969, Amer. Mus. Novit., 2398: 6, 28 (no
locality).

Hemisorubim platyrrhynchus ; A. de Miranda Ribeiro, 1914, Comm. Linh. Telegr. Estr.
Matto-Grosso ao Amazonas, 15, Annexo 5: 7 (Manaos).

Hemisorubim platyrhinchos ; Mago-Leccia, 1967, Soc. Venez. Cienc. Nat. Bol., 27: 255
(Venezuela).

Material. — One specimen, 29 December 1963, Sarakreek near Locus (Boeseman,
RMNH no. 25767), standard length 112 mm. One specimen, 7 February 1964, Suriname
Rivier near Brokopondo (Boeseman, RMNH no. 26521), standard length 151 mm. One
specimen, 27 February 1964, Suriname Rivier near Brokopondo (Boeseman, RMNH
no. 26522), standard length 314 mm. One specimen, 20 March 1965, Suriname Rivier
near Brokopondo (Mees, RMNH no. 26586), standard length 420 mm. One specimen,
20 March 1965, Suriname Rivier near Brokopondo (Mees, RMNH no. 26587), standard
length 341 mm. One specimen, 20 March 1965, Suriname Rivier near Brokopondo (Mees,
RMNH no. 26846), standard length 225 mm. One specimen, 15 February 1967, Broko¬
pondo Meer near Affobakka (Nijssen, ZMA no. 106845), standard length 190 mm.

Characters. — D 1 . 6 , A 10-n (iii or iv.7), P I.9, V 6 (1.5), C 16-18
(branched rays only), branchiostegals 10-11, gill-rakers 2 4 - 1 +5 to 7.
Further characters are given in the generic diagnosis.

Colours highly diagnostic. Body light brownish, whitish below, with on
each side from three to five moderately large roundish black dots, the first
one usually just behind the head, the last one on the caudal base.

Distribution. — The species ranges from the Orinoco basin through
eastern South America to the Parana-Paraguay basin. In Suriname we have

120

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

only obtained it in the Suriname Rivier near Brokopondo and in its tributary
the Sara Kreek, but it is probably widely distributed.

Habitat — As far as can be said from our few specimens, this large
species is confined to the deeper parts of the large rivers. This habitat pre¬
ference probably accounts for the fact that we failed to secure more specimens
(remarks made in the discussion of Pseudo platy stoma fasciatum would also
apply to the present species), but the species appears to be uncommon.

Genus Brachyplatystoma Bleeker

Brachyplatystoma Bleeker, 1862, Atlas Ichth., 2: 10 — type by original designation,
Platystoma Vaillantii Valenciennes.

Piratinga Bleeker, 1862, Atlas Ichth., 2: 11 — type by original designation, Bagrus
reticulatus Kner.

Malacobagrus Bleeker, 1862, Atlas Ichth., 2: n — type by original designation, Pime-
lodes filamentosus Lichtenstein.

Generic diagnosis. — Large species with strongly depressed snout, but the
profile ascending and the posterior part of the head less depressed: its depth
at the level of the cleithra 0.6-0.75 times its width at the same place; eyes
small, dorsolateral in position, over five times in snout, over eleven times in
head, with a free rim; snout depressed, broadly rounded, upper jaw pro¬
truding; teeth in a band in upper and lower jaw, those in the upper jaw
slightly larger than is usual in the family, not contiguous with a band of
smaller teeth covering vomer and palatines (fig. 32b); postoccipital process
long, comparatively narrow, of even width throughout, ending in a deep fork,
into which fits the slender tip of the predorsal plate, these bones, however,
are not fused, but remain narrowly separated; postoccipital process almost
twice the length of the predorsal plate 1 ); dorsal and pectoral spines not
encased in skin, well-developed but not very strong; maxillary barbels very
long, reaching to caudal peduncle or much beyond, sometimes over twice the
body length in B. filamentosum but much shorter in some other species and
apparently in very large specimens, of sizes which have not been available;
no pectoral pore; caudal fin in all species deeply forked, the outer rays
filamentous, sometimes very long; colour plain or with indistinct cross-bars,
or with darfc dots on the flanks.

Distribution. — Eastern Colombia, Venezuela, the Guianas, Amazonia

1) In B. flavicans the postoccipital process has been described as triangular, short,
not reaching the predorsal plate and in B. juruense as slender but not reaching the pre¬
dorsal plate. The Eigenmanns (1890: 196) observed that in small specimens of B. vail¬
lantii the postoccipital process reaches to the predorsal plate, but that in large specimens
it does not. Apparently the shape and length of the occipital process is not a good specific
character in the genus Brachyplatystoma.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

121

inland to Peru and Bolivia, the Rio Parnaiba, and eastern Brazil, south to the
Rio Paraiba in south-eastern Brazil (cf. Eigenmann & Eigenmann, 1890:
i 97 )-

Discussion. — The genus is, like so many genera of the two families here
dealt with, in need of a critical revision. It appears reasonably certain that
there are more names available than valid species, but in the most recent
lists (Gosline, 1945; Fowler, 1951) the allocation of names appears to have
been an arbitrary process, not based on personal study. Gosline admits seven
species, with synonyms as follows: B. filamentosum (Lichtenstein) (syn.
Bagrus reticulatus Kner, Piratinga pira-aiba Goeldi, Platystoma verrucosum
Boulenger, Bagrus rousseauxii Castelnau), B . vaillanti (Cuvier & Valen¬
ciennes) (syn. B. parnahybae Steindachner), B. flavicans Castelnau (syn.:
Bagrus goliath Kner), B. paraense Steindachner, B. affine (Cuvier & Valen¬
ciennes), B. juruense Boulenger, and B. parnahybae Steindachner (syn.: B.
affine Steindachner, nec Cuvier & Valenciennes). Note that Gosline has
listed the name B. parnahybae twice, once as a synonym of B. vaillantii , once
as a valid species.

Fowler (1951) recognizes eight species: B. affine (Valenciennes), B.
filamentosum (Lichtenstein) (syn.: Bagrus rousseauxii Castelnau, Bagrus
reticulatus Kner, Platystoma mucosa Vaillant, Platystoma verrucosum Bou¬
lenger, Piratinga pira-aiba Goeldi), B. flavicans (Castelnau) (syn.: Bagrus
goliath Kner), B. goeldii Eigenmann & Bean, B . juruense (Boulenger), B.
paraense Steindachner, and B . vaillantii (Valenciennes). The difference
from Gosline is that Fowler has definitely accepted B. parnahybae , and has
added B. goeldii , a name overlooked by Gosline.

I have not attempted to sort out all these names, mainly because many
of the type-specimen's are very large, for which reason I have been reluctant
to ask for loans, but I have followed Bertin & Esteve (1950), who placed
B. affine in the synonymy of B. vaillantii , as, indeed, Eigenmann & Eigen¬
mann (1890: 196) had done many years earlier, whereas on the following
pages evidence will be presented that B. goeldii is another synonym of B.
filamentosum . This leaves the genus with six nominal species.

Two species occur in Suriname; these species can easily be told apart
although they have been confused by van der Stigchel (1946, 1947) whose
description was based on mixed samples.

Brachyplatystoma filamentosum (Lichtenstein)

Pimelodes filamentosus Lichtenstein, 1819, Zool. Mag. (Wiedemann), 1 (3) : 60 —
Siidamerika.

Bagrus rousseauxii Castelnau, 1855, Anim. Nouv. l’Amer. Sud, Poiss.: 32, pi. 14
fig. I — la riviere des Amazones.

122

ZOOLOGISCHE VERHANDELINGEN 13 2 (l 974 )

B[agrus] reticulatus Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw. Cl., 26
(1857) : 376 — Salto Theotonio am Flusse Araguay, Forte do Rio branco und Rio
Madeira.

Platystoma mucosa Vaillant, 1880, Bull. Soc. Philomat., (7) 4: 151 — Calderon (Haute-
Amazone).

Platystoma verrucosum Boulenger, 1881 (?), Zool. Rec. Pisces, (1880): 19 — lapsus
only for P. mucosa Vaillant.

Pirating a pird-aiba Goeldi, 1898, Bol. Mus. Paraense, 2: 467, 477, pi. Novos Peixes
de Amazonia, fig. 4 — Para.

Br achy platystoma goeldii Eigenmann & Bean, 1907, Proc. U.S. Nat. Mus., 31 : 661,
fig. 1 — Amazon River between Para and Manaos.

Piratinga reticulata ; Bleeker, 1862, Atlas Ichth., 2: 11 (name only); Bleeker, 1863,
Ned. Tijdschr. Dierk., 1: 99 (name only) ; Gunther, 1864, Cat Fish. Brit. Mus., 5:
113 (Salto Theotonio on the River Araguay, Forte de Rio Branco, Rio Madeira).

Malacobagrus filament osus ; Bleeker, 1862, Atlas Ichth., 2: n (name only); Bleeker,
1863, Ned. Tijdschr. Dierk., 1 : 100 (name only).

Piratinga filamentosa ; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 112 (Brazil); Goeldi,
1901, Bol. Mus. Paraense, 3: 181-194, pi. 1, 2 (Amazonas).

Brachyplatystoma filamentosum ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad.
Sci., (2) 1: 136 (name only) ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad.
Sci., 1: 195 (discussion of generic allocation; no material); Eigenmann, 1910, Rep.
Princeton Univ. Exp. Patagonia, 3: 390 (Brazil; Para); A. de Miranda Ribeiro, 1911,
Arch. Mus. Nac. Rio de Janeiro, 16: 329 (Araguaya, Salto Theotonio, Forte do Rio
Branco, Rio Madeira, Tocantins, Amazonas); Starks, 1913, Stanford Publ. Univ. Ser.:
27 (Para); A. de Miranda Ribeiro, 1914, Comm. Linh. Telegr. Estr. Matto-Grosso ao
Amazonas, 15, Annexo 5: 8 (Manaos); A. de Miranda Ribeiro, 1920, Comm. Linh. Telegr.
Estr. Matto-Grosso ao Amazonas, 58, Annexo 5: 12 (Aripuanan) ; Gosline, 1945, Bol.
Mus. Nac. Rio de Janeiro, Zool., 33: 49 (Brasil); Fowler, 1951, Arq. Zool. S. Paulo, 6:
584 (Amazonia, Rio Parnaiba, Guiana, Venezuela).

Brachyplatystoma reticulata ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) I : 136 (name only).

Brachyplatystoma rousseauxii ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad.
Sci., (2) 1 : 136 (Para) ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1 :
198 (Para); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 31 (Amazon);
Eigenmann & Bean, 1907, Proc. U.S. Nat. Mus., 31: 662 (Amazon River between Para
and Manaos); Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 390 (Amazon
and its tributaries); A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16:
327, fig. 128 (Amazonas e tributaries) ; Starks, 1913, Stanford Publ. Univ. Ser.: 27
(Para).

Brachyplatystoma reticulatum ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad.
Sci., 1 : 198 (Rio Tocantins, Amazon and its tributaries; Rio Madeira) ; Eigenmann &
Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 31 (Rio Tocantins; Amazon and tributaries;
Rio Madeira) ; Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 390 (Rio
Tocantins; Amazon; Rio Madeira).

Piratinga Rousseauxii ; Goeldi, 1898, Bol. Mus. Paraense, 2: 457, 477 (Magoary).
Brachyplatystoma rousseauxi ; A. de Miranda Ribeiro, 1914, Comm. Linh. Telegr.
Estr. Matto-Grosso ao Amazonas, Annexo 5: 8 (Manaos) ; Fisher, 1917, Ann. Carnegie
Mus., 11 : 418 (Para).

Brachyplatystoma goeldii ; Eigenmann & Allen, 1942, Fish. W. South America: 112
(Iquitos, Peruvian Amazon); Fowler, 1945, Los Peces del Peru: 50 (Peru: Iquitos) ;
Fowler, 1951, Arq. Zool. S. Paulo, 6: 585 (Amazonia, Peru).

Brachyplatystoma vaillanti; (pt.) van der Stigchel, 1946, South American Nemato-
gnathi: 65 (Surinam); (pt.) van der Stigchel, 1947, Zool. Meded., 27: 65 (Surinam).

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

123

Material. One specimen, before 1819, Siidamerika (Wustlandt and Graf von Hoff-
mannsegg, ZMB no. 2973), standard length ca. 165 mm, holotype of the species. One
specimen, June 1911, Suriname (Jhr. W. C. van Heurn, RMNH no. 17291), standard
length 405 mm. Two specimens, summer 1911, Suriname (Jhr. W. C. van Heurn, RMNH
no. 11060), standard length 224, 229 mm. Two specimens, 1911, Paramaribo, Suriname
(Jhr. W. C. van Heurn, RMNH, removed from sample no. 17324 and re-registered under
no. 26585), standard length 213, 226 mm. One specimen, 6 July 1961, Georgetown fish-
market, British Guiana (R. H. McConnell, BM no. 1971.7.29: 95), standard length
260 mm. One specimen, 27 June 1966, mouth of Coppename and Saramacca Rivier (W.
Vervoort, RMNH no. 25650), standard length 176 mm.

Characters. — D 1 . 6 , A 12-13, P 1-8 or I.9, V 6 (i.5), C i.ig.i to i,i7-i and
rudiments, branchiostegals 11, gill-rakers 17-21.

Upper jaw strongly protruding over lower jaw; depth of head at level of
cleithra 0.6-0.7 times its width at the same place; maxillary barbels excessively
long, more than twice the standard length 1 ); outer pair of mental bai^els
comparatively short, reaching to pectoral base or at most to half-way along
pectoral spine; gill-rakers shorter and less numerous than in the following
species, 17-21 on outer branchial arch; pectoral fins with one spine and 8
or 9 rays; ventral fins distally more or less rounded with little difference in
length between the first and the last ray; adipose fin comparatively short,
its base opposite of, and of the same length as, the anal base, 7.2-8.2 times
in standard length.

Colours of preserved material: dull brown above the lateral line, whitish
below, on the 'body above the lateral line some eight vague roundish dark
dots, the last one on the caudal peduncle, near the base of the upper lobe of
the caudal fin. The Georgetown specimen is plain, without dots; as the dots
are still definitely present in material collected sixty years ago and the
Georgetown fish is otherwise well-pigmented, it is likely that it has never
had spots; more material must show if this is an individual aberration, or
perhaps a permanent character of specimens from British Guiana; I note
that the large fish from Para figured in colour by Goeldi (1901) is also plain.

Distribution. — Widely distributed in Amazonia where, however, only
known from the larger rivers, upstream to Iquitos. Here recorded for the

1) It is generally known that large specimens (the species can attain a length of two
metres) have relatively much shorter maxillary barbels. An excellent illustrated account
of differences between specimens of various sizes was given by Goeldi (1901), to
whose paper I refer for further information on this subject. No large specimens have
been available to me and in the largest one examined (RMNH no. 17291) the maxillary
barbels are broken; even so the longest barbel reaches to beyond the anal origin, and
would have amply exceeded the standard length, but might not have attained twice the
standard length. In the specimen from British Guiana the right maxillary barbel is over
twice the standard length, but the left one, which appears to be complete and not
obviously damaged, only ca. 1.5 times the standard length.

124

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

first time from the Guianas (Suriname and Demerara), where confined to
the lower courses and mouths of the large rivers. No apparent ecological
difference from B. vaillantii. I have failed to find on what Fowler's (1951)
inclusion of Guiana and Venezuela into the range of this species was based,
and presume that he made an error, although now it has been found in the
Guianas, and may conceivably occur in Venezuela. As mentioned on a later
page I believe Puyo’s B. rousseauxii to be referable to B. vaillantii.

Discussion. — Lichtenstein’s (1819) description of Pimelodes filamentosus
is by modern standards inadequate, and I am not aware that the type had
ever been re-examined, so that its status remained somewhat doubtful. For¬
tunately the type still exists in Berlin, and was made available to me. Although
it is in a very poor condition it still shows the following diagnostic characters,
leaving no doubt about its identity: D 1 . 6 , A 12, P I.9, gill-rakers 18 (6 + I
+ 11), ventrals rounded, upper jaw strongly protruding, maxillary barbels
very long, adipose fin 7.4 times in standard length.

Almost immediately after its description, Eigenmann & Eigenmann
(1888: 136) relegated Platystoma mucosa to the synonymy of Brachyplaty -
stoma vaillantii but certainly in error, for in the description of P. mucosa
several characters are given which show that it belongs without question to
the present species: the long barbels (more than twice the standard length),
the number of rays in the pectoral fin ( 1 . 8 ), and the coloration: “La colora¬
tion sur I'animal conserve dans la liqueur parait uniforme; on distingue une
ou deux taches noires sur les flanes au niveau de la dorsale, et une tache
arrondie, petite, plus nette, de meme couleur, a l’extremite du pedoncule
caudal” (Vaillant, 1880). Later authors, like Gosline (1945) and finally Bertin
& Esteve (1950), who examined the type-specimen, have correctly syno-
nymized P. mucosa with B. filamentosum.

The synonymy of Bagrus rousseauxii Castelnau with B. filamentosum has
been confirmed by Bertin & Esteve (1950: 58), who examined its type-
specimen. Only four years after its description, A. de Miranda Ribeiro
(1911 : 326) questioned the validity of B. goeldii, suggesting that it was only
a variety of B . vaillantii . Of later authors, van der Stigchel (1946) placed
B. goeldii in the synonymy of B. vaillantii, Fowler (1945b, 1951) accepted it
as valid, and Gosline (1945) ignored the name altogether. Although one may
assume that Eigenmann & Allen (1942), who listed material, regarded B.
goeldii as a valid species, they did not discuss it in any way, so that the latest
published opinion on its status is that of A. de Miranda Ribeiro referred to
above. The rather scanty description and the good figure of B. goeldii
published by Eigenmann & Bean (1907) agree perfectly with my material
of B. filamentosum.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

125

Brachyplatystoma vaillantii (Valenciennes) (fig. 32b, pi. 15)

Platystoma Vaillantii Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.), 15: 16, pi. 423 — de Cayenne et de Surinam.

Platy stoma affine Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.), 15: 18 — no locality — Rio de Janeiro (cf. Bertin & Esteve, 1950: 57).

Platy stoma affine ; Castelnau, 1855, Anim. Nouv. l’Amer. Sud. Poiss.: 40 (l'Araguay);
Gunther, 1864, Cat. Fish. Brit. Mus., 5: 109 (no locality).

Platystoma Vaillantii ; Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw. CL, 26
(1857): 397, pi. V fig. 10 (Para); Peters, 1877, Mber. Akad. Wiss. Berlin: 469 (Cala-
bozo, Venezuela); Goeldi, 1898, Bol. Mus. Paraense, 2: 464, 476 (Magoary; Counany).

Brachyplatystoma Vaillanti ; Bleeker, 1862, Atlas Ichth., 2: 10 (name only); Bleeker,
1863, Ned. Tijdschr. Dierk., 1 : 97 (name only); Bleeker, 1864, Nat. Verh. Holl. Maat-
sch. Wetensch., (2) 20: 70 (Surinama).

Platystoma vaillantii ; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 108 (Demerara).

Brachyplatystoma vaillanti ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) 1: 136 (Tabatinga; Para; Porto do Moz; Avary; Rio Puty; Juiz de Fora on the
Parahyba); Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1: 196 (Taba¬
tinga; Para; Porto do Moz; Avary; Rio Puty; Juiz de Fora, on the Parahyba; Serpa,
Rio Negro); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 30 (Eastern
slopes of South America north of Rio Parahyba); Eigenmann & Bean, 1907, Proc. U.S.
Nat. Mus., 31: 662 (Amazon River between Para and Manaos); Eigenmann, 1910, Rep.
Princeton Univ. Exp. Patagonia, 3: 390 (Eastern slopes north of Rio Parahyba) ; A.
de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 323, fig. 124 (distribu¬
tion) ; Eigenmann, 1912, Mem. Carnegie Mus., 5: 180 (British Guiana: Rockstone;
Georgetown market); Starks, 1913, Stanford Publ. Univ. Ser.: 27 (Para); A. de
Miranda Ribeiro, 1914, Comm. Linh. Telegr. Estr. Matto-Grosso ao Amazonas, 15,
Annexo 5: 8 (Manaos) ; Steindachner, 1915, Denkschr. Akad. Wiss. Wien, Mathem.-
Naturw. Kl., 93: 64 (Para); Fisher, 1917, Ann. Carnegie Mus., 11: 418 (Para; Santa-
rem; Manaos); Delsman, 1941, Verh. Kon. Natuurh. Mus. Belgie, (2) 21 (3): 79
(Trombetas-river, near Obidos; Manaos (in the harbour)); Schultz, 1944. Proc. U.S.
Nat. Mus., 94: 232 (Venezuela); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Z00L,
33: 49 (Amazonas; Pamaiba); (pt.) van der Stigchel, 1946, South American Nemato-
gnathi: 65 (Surinam; Para, Brazil; Georgetown market, British Guiana); (pt.) van
der Stigchel, 1947, Zool. Meded., 27: 65 (Surinam; Para, Brazil; Georgetown market,
British Guiana) ; Puyo, 1949, Poiss. Guyane Fr.: 85-87, fig. 48 (Guyane Frangaise) ;
Bertin & Esteve, 1950, Cat. Types Poiss. Mus. Paris, 5: 57 (FI. Surinam ou Cayenne
(Guyane); FI. Surinam (Guyane); Rio de Janeiro (Bresil)); Lundberg & Baskin,
1969, Amer. Mus. Novit., 2398: 6, 28 (no locality).

Brachyplatystoma vaillianti; Beebe, 1925, Zoologica, 6: 121 (Kartabo).

Brachyplatysoma vaillanti ; Roth, 1943, Fish Life in British Guiana: 17 (British
Guiana).

Hypophthalmns Dawalla , or Piratinga Reticulata , Brachy platy soma vaillante ; Roth,
1943, Fish Life in British Guiana: 74 (British Guiana: Mazaruni above Kapashi Island).

Brachyplatystoma rousseauxii ; (?) Puyo, 1949, Poiss. Guyane Fr.: 82-85, fig. 45, 46
(Guyane Frangaise).

Brachyplatystoma vaillantii ; Fowler, 1951, Arq. Zool. S. Paulo, 6: 586 (Amazonas,
Rio Parnaiba, Guiana, Venezuela); Dahl, 1961, Noved. Colomb., x: 509 (Guayabero
River, Colombia).

Material. — One specimen, 1908, Georgetown market, British Guiana (C. H. Eigen¬
mann, ZMA no. 110702), standard length 169 mm. Two specimens, 1911, Paramaribo,
Suriname (Jhr. W. C. van Heum, RMNH no. 17324), standard length 186, 188 mm.
One specimen, 6 July 1961, Bourda fish market, Georgetown (R. H. McConnell, BM

126

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

no. 1971. 7.29: 96), standard length 185 mm. One specimen, 10 September 1961, W. bank
Demerara River, Georgetown (R. H. McConnell, BM no. 1971.7.29: 97), standard length
167 mm. Five specimens, 12 September 1961, Bourda fish market, Georgetown (R. H.
McConnell, BM no. 1971.7.29: 98-102), standard length 148, 162, 170, 188, 197 mm.
Twelve specimens, 27 October 1961, Bourda fish market, Georgetown (R. H. McConnell,
BM no. 1972.7.27: 679-690), standard length 125-152 mm. One specimen, 22 June 1966,
mouth of Coppename near Coppename Punt, Suriname (Nijssen, ZMA no. 106843),
standard length 242 mm. One specimen, 27 June 1966, mouth of Coppename and Sara-
macca Rivier, Suriname (W. Vervoort, RMNH no. 26590), standard length 284 mm.
Our collection further contains some old specimens from Suriname (leg. Dieperink) and
elsewhere, which were correctly identified by van der Stigchel, and therefore have not
been re-studied (RMNH nos. 3016, 3017, 3018, 14762, 16051).

Characters. — D I. 6, A 13-15 (iv or V.9J4 to n), P I.io or I.n, V 6
(1.5), C i.i5.i and rudiments, branchiostegals 11, gill-rakers 25-31.

Upper jaw only slightly protruding over lower jaw; depth of head at level
of deithra 0.7-0.75 times width at the same place; maxillary barbels very
long, reaching to the caudal tip; outer pair of mental barbels long, reaching
to beyond ventral origin or even to beyond anal origin; gill-rakers numerous,
25-31 on outer branchial arch; pectoral fins with one spine and 10 or 11 rays;
ventral fins pointed, the inner ray no more than 0.6 times the length of the
outer ray; adipose fin comparatively long, its base distinctly longer than the
anal base, 4.25-4.8 in standard length.

Colours of preserved material: dorsal surface plain light brown, without
any markings, ventral surface whitish.

Size. Dahl (1961: 509) reported from Colombia: 'The biggest specimen
seen by the author had a weight of about 90 kilogrammes, but it is told that
specimens up to and even above 200 kilogrammes are not rare in the lower
parts of the Guaviare River. Shark hooks and manila rope are sometimes
used to catch the larger ones. Professor Medem on one occasion saw a
rather big shark hook that had been almost completely straightened by a
giant “valenton” Such monsters have not, to my knowledge, been recorded
in Suriname 1 ).

Distribution. — Tropical South America east of the Andes, from Colombia
(Guayabero River) and Venezuela (Calabozo), through the Guianas and
Amazonia to the Parahyba. Although in the Amazon basin the species has
been found upstream to Tabatinga, in Suriname it appears to be confined to

1) From Dr. Boeseman I received a colour slide of a specimen of about 170 cm total
length, captured at Brokopondo in February 1964, before the dam was closed. This
proves that large individuals do occur in Suriname and that, at least in the larger rivers,
the species may be found some distance upstream. The picture shows the fish with blue
(near to Neropalim Blue of Ridgway, 1912: pi. xxii) upper parts, white below. A black-
and white photograph, made from the slide, although not quite sharp, appears to be of
sufficient interest to be reproduced here (plate 15).

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

127

the river mouths, the only certain localities being Paramaribo (Suriname
Rivier), and the Coppename-Saramacca mouth. The same is true for French
Guiana (Puyo, 1949: 87), and for British Guiana, for although Eigenmann
(1912: 180) mentioned a specimen from Rockstone (“head only”), all other
records are from coastal areas and there is nothing to support Eigenmann’s
(1912: 76 footnote 34) contention that: “This species certainly occurs
generally inland’.

Discussion. — Puyo’s (1949) description and figures of “Brachyplaty-
stoma rousseauxii” appear to be referable to large individuals of B. vaillantii
rather than to B . filamentosum (cf. adipose base longer than anal base, upper
jaw only a little longer than lower jaw, no spots on the body).

Genus Pseudoplatystoma Bleeker

Pseudoplatystoma Bleeker, 1862, Atlas Ichth., 2: 10 — type by original designation,
Silurus fasciatus Bloch = Silurus fasciatus Linnaeus.

Hemiplatystoma Bleeker, 1862, Atlas Ichth., 2: 10 — type by original designation,
Platystoma tigrinum Valenciennes.

Generic diagnosis. — Very large species with big, moderately depressed
head; width between the cleithra about 1.3 times depth at the same place;
eyes comparatively small, not quite dorsal in position, with a free rim; snout
long, depressed, upper jaw protruding; teeth smallish, in a band in each jaw
and patches on vomer and palatines (fig. 32c); postoccipital process well-
developed, much longer than wide, meeting the slender predorsal plate; dorsal
and pectoral spines well-developed, but encased in thick skin and therefore not
conspicuous; maxillary barbels not very long, varying from reaching to half¬
way along the pectorals, to just reaching anal origin. Further characters have
been given in the key.

Distribution. — Northern and eastern South America, east of the Andes,
south to the Parana-La Plata basin.

Discussion. — Three species are currently recognised: P. fasciatum (Lin¬
naeus), P. coruscans (Agassiz) and P. tigrinum (Valenciennes). Besides,
P. fasciatum has been divided into a number of subspecies, about which more
will be said in the discussion of that species. The validity of P. tigrinum
requires confirmation, it may be a synonym of P. fasciatum.

Pseudoplatystoma fasciatum (Linnaeus) (fig. 32c)

[Silurus] fasciatus Linnaeus, 1766, Syst. Nat. (ed. 12), 1: 505 — Habitat in Brasilia,
Surinami, coenis expetita.

Silurus fasciatus] Gmelin, 1789, Syst. Nat. (ed. 13), 1 (3): 1359 (Habitat Surinami,
in Brasilia); Bloch, 1797, Ichth. Hist. Nat. Poiss., 11: 22, pi. 366 (Surinam).

Silurus Fasciatus ; Bloch & Schneider, 1801, Syst. Ichth,: 382 (Habitat in Surinamo).

128

ZOOLOGISCHE VERHANDELINGEN 13 2 (1974)

Pimelodus fasciatus ; La Cepede, 1803, Hist. Nat. Poiss., 5: 94, 99-101 (Surinam).

Platystoma fasciaium; Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat.
Poiss. (4 0 ed.), 15: 11 (Surinam); Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-
Naturw. Cl., 26 (1857): 401 (Surinam); Gunther, 1864, Cat. Fish. Brit. Mus., 5: 107
(Essequibo; Surinam; River Capin, Para); Kappler, 1885, Das Ausland, 58: 899 (Hol-
landisch Guiana; Armina Fall); Kappler, 1887, Surinam: 150 (Surinam; Arminafall).

Pseudoplatystoma fasciatnm ; Bleeker, 1862, Atlas Ichth., 2: 10 (name only) ; Bleeker,
1863, Ned. Tijdschr. Dierk., 1 : 97 (name only); Bleeker, 1864, Nat. Verh. Holl. Maat-
sch. Wetensch., (2) 20: 72 (Surinama); Eigenmann, 1912, Mem. Carnegie Mus., 5: 182
(British Guiana: Rupununi Pan, Wismar) ; Popta in Benjamins & Snelleman, 1914-1917,
Encycl. Ned. West-Indie: 592 (Suriname); van der Stigchel, 1946, South American
Nematognathi: 68 (Surinam); van der Stigchel, 1947, Zool. Meded., 27: 68 (Surinam);
Boeseman, 1953, Zool. Meded., 32: 7 (Leonsberg); Lowe, 1964, J. Linn. Soc. Lond., Zool.,
45: no, III, 114, 119, 141, 143 (Rupununi district); McConnell, 1967, Timehri, 43: 67
(Rupununi).

Pseudoplatystoma fasciatus; McConnell, 1967, Timehri, 43: 65 (Pirara creek,
Rupununi).

pseudoplatystoma fasciatnm; Nijssen, 1970, Schakels, S 76: 3, 4, pi. (Suriname).

Note. — In view of the uncertainty concerning the validity of subspecies,
the bibliography is limited to references concerning occurrence in the Guianas.

Material. — One specimen, 20 March 1965, Suriname Rivier near Brokopondo (Mees,
RMNH no. 26588), standard length 715 mm. One specimen, head only, 20 February
1966, Sipaliwini (Mees, RMNH 26589), head slightly smaller than that of the first
specimen. In addition there are 19 specimens, mostly in the size-range 45-60 cm standard
length, in ZMA (leg. Nijssen), which I have not studied.

Characters. — D 1 . 6 , A 14 (v.9), P I.io, V 6 (i.5), C i.i5.i and rudiments,
branchiostegals 14, gill-rakers less than 10 (1 + 1 + 1 or 2 and several
rudimentary ones).

Distribution and habitat. — The species has been recorded from a very
wide area in eastern South America, but until its synonymy has been sorted
out, there remains an element of uncertainty in many records. In Suriname
it appears to be widely distributed in the large rivers, but also in deeper parts
of the smaller rivers quite far upstream, as proved by the example from
Sipaliwini and one I saw in the upper course of the Sara Kreek. The popu¬
lation density of a species of this size is presumably much less than that of
the smaller species, and it would also be more difficult to catch. There is no
reason, however, to believe that it is uncommon, and Kappler (1885) described
it as: “im inneren Lande sehr hauftig”. Lowe (1964) found it common in the
Rupununi District of British Guiana.

Discussion. — Next to Brachyplatystoma, this is the largest of the fresh¬
water inhabiting Nematognathi known from Suriname. The tiger-stripes on
the body further make it easy to identify.

The Eigenmanns (1888, 1889) have described and named several varieties,

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

129

based mainly on length of barbels and differences in markings. These
varieties have later become accepted as subspecies (Eigenmann & Eigen-
mann 1890; Gosline, 1945; Fowler, 1951).

Pseudo platystoma fas datum being such a large species, captured individuals
are more likely to end up in the frying pan or in the cooking pot than in the
collector’s bin, and the various supposed subspecies remain known from very
few individuals, so that the possibility that some of them are based on indi¬
vidual variants cannot be excluded. Anyway, these subspecies have never
been considered in the light of the modern subspecies-concept, according to
which a subspecies is a geographical race; consequently (excluding exceptions
which to this general discussion are irrelevant) no two subspecies can occur
together. However, P. fasdatum intermedium was described from Obidos,
whence the Eigenmanns also recorded specimens of the nominate race.
Indeed, the distribution of the nominate race as given in recent literature,
includes the ranges of all the subspecies, an impossible situation.

When a species is divided into geographical races, the first obvious thing
to do is restrict its type-locality, so that one knows what the nominate race
is, but even this has never been done. Linnaeus (1766: 505) gave: “Habitat
in Brasilia, Surinami, coenis expetitia”, and his references show that the
locality Suriname was based on Seba (1758: 84, pi. 29 fig. 67), whereas the
reference to Brasil was derived, through Seba, from Marcgrav. I here restrict
the type-locality to Suriname, which makes my Brokopondo specimen topo-
typical.

Until the position of the various subspecies has been clarified, there does
not appear to be any point in using a trinomial. A study of P. fasdatum
should also include P. tigrinum, which appears very doubtfully distinct, and
several names given to individuals from Amazonia, which have been placed
in the synonymy of nominate P . fasdatum , and have never been re-evaluated
when subspecies were later described from the same region.

P . fasdatum is a fish of excellent edible quality; the flesh is firm, and best
when fried or grilled. As mentioned above, this fact, together with a size
which makes conservation difficult, has certainly contributed to its being
somewhat scarce in collections.

Van der Stigchel (1946: 68, 69) states in the generic diagnosis of Pseudo-
platystoma “bafbels short”, but in the description of the species P. fasdatum ,
the only member of the genus which he examined : “maxillary barbels reaching
the tip of the caudal”. I have checked this character in the three specimens
in Leiden on which his description was based and found that all three have
the maxillaiy barbels just reaching the base of the anal fin.

9

I 3 °

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Genus Pimelodus La Cepede

Pimelodus La Cepede, 1803, Hist. Nat. Poiss., 5: 93 — based on a whole series of
species, now placed in several families; type by subsequent selection, Pimelodus maculatus
La Cepede.

Generic diagnosis. — D 1 . 6 , A 10-14, P I.9 or 1 .10, V 6, C i.iS-i and rudi¬
ments. Sturdy fishes of moderate to fairly large size; skull and nuchal bones
covered with thin skin only; postoccipital process with a broad base, tapering
backwards to meet the small dorsal plate; teeth in a band in each jaw, none
on vomer and palatines (some extra-limital species have teeth on vomer
and pterygoids); barbels long, maxillary barbels usually reaching to caudal
base or beyond (in P. ortmanni the maxillary barbels reach only just beyond
the ventral base); a minute pore above the posterior part of the pectoral base
( P . blochii, P . albofasciatus) , or no pore (P. ornatus ); dorsal and pectoral
spines strong, pungent, more or less serrated; adipose fin well-developed;
caudal fin large, forked.

In the species occurring in Suriname, the adipose fin is short and high,
almost triangular in shape, but some members of the genus, in particular
P. altissimus, have a long adipose fin.

Discussion. — The genus is widely distributed in tropical South America,
from Panama to the Rio de la Plata. Over twenty species have been described,
some of which are of doubtful validity.

In Suriname, the genus is represented by three species: P. ornatus is a
well-differentiated species with a wide distribution in South America, and
unlikely to be confused with any other species. The other two species, P.
blochii and P. albofasciatus are not only extremely similar to each other, but
belong to a group of confusingly similar forms which for purposes of dis¬
cussion may be referred to as the P. maculatus-group after the first-described
species. A revision of this group, however desirable, goes beyond the scope
of this contribution, and could, I believe, only be carried out successfully on the
basis of systematic collecting of fresh material in all parts of its extensive
range (under the names P. maculatus, P. clarias, etc., members of the group
have been recorded from Uruguay and Bolivia to Venezuela and Panama) 1 ).
How many described forms belong to the group is not certain; of the species
and subspecies currently recognized at least the following can be placed in

1) Such a study would certainly be very rewarding, and is also urgent, as in Brazil
experiments of transplantation of Pimelodus “ clarias ” have begun (cf. Schubart, 1964:
19). If these ill-advised experiments are “successful”, they will not fail to foul up the
zoogeography of this interesting group of fishes. These introductions appear particularly
unnecessary, if at the places of introduction very closely related forms do occur
naturally.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE 131

it: P. maculatus La Cepede, 1803, P. argenteus Perugia, 1891, P. ortmanni
Haseman, 1911, P. blochii Valenciennes, 1840, P. coprophagus Schultz, 1944,
P. punctatus (Meek & Hildebrand, 1913), P. albofasciatus n. sp. and probably
P. garcia-barrigai Dahl, 1961. I consider it likely that varieties or forms as
described by Eigenmann (1912) and Devincenzi & Teague (1942 : 44-45) are
also at least valid subspecies. Anyway, the new information gained from the
material collected in Suriname is that two forms which formerly would have
been dismissed as varieties, are morphologically and ecologically differ¬
entiated and apparently geographically separated. Therefore, they should
be recognised nomenclaturally, either as subspecies or as species. The reason
why I am treating both forms from Suriname as species (and not for example
as subspecies of P. maculatus ), is because the arbitrary application of trino¬
mials to some forms is meaningless, and quite likely erroneous, until the
whole group has been revised.

Key to the species of Pimelodus known from Suriname

ia — A large black blotch present on the dorsal fin; gill-rakers on outer

branchial arch 14-21 (usually 15-18).P. ornatus

l b —■ Dorsal fin without markings; gill-rakers on outer branchial arch

21-30 (usually 23-27).2

2a — Colour plain, light liver-brown; eye smaller, 1.0-1.7 m b° n y inter¬
orbital ...P. blochii

b — Colour dark greyish with a conspicuous broad white longitudinal band
along the sides; eye larger, 0.7-1.0 in bony interorbital P. albofasciatus

Pimelodus ornatus Kner

Pimel[odus] ornatus Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw. Cl.,
26 (1857) : 4n, pi. VI fig. 18 — von Surinam, dem Rio Negro und Cujaba.

Megalonema rhabdostigma Fowler, 1914 Proc. Acad. Nat. Sci. Philad., 66: 256,
fig. 10 — Rupununi River, British Guiana.

Pimelodus ornatus ; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 116 (River Capin, Para);
Peters, 1877, Mber. Akad. Wiss. Berlin: 470 (Venezuela: Calabozo); Kappler, 1881,
Hollandisch-Guiana: 167 (Hollandisch-Guiana); Kappler, 1885, Das Ausland, 58: 919
(Hollandisch Guiana); Kappler, 1887, Surinam: 152 (Surinam); Eigenmann & Eigen¬
mann, i 838 , Proc. Calif. Acad. Sci., (2) 1 : 134 (Goyaz) ; Eigenmann & Eigenmann,
1891, Proc. U.S. Nat. Mus., 14: 29 (Amazon, Solimoens, and northward); Goeldi, 1898,
Bol. Mus. Paraense, 2: 456, 461 (Rio Negro, ex Kner; Igarape-mirim); Regan, 1905,
Proc. Zool. Soc. Lond.: 190 (Rio Negro) ; Eigenmann, 1910, Rep. Princeton Univ. Exp.
Patagonia, 3: 388 (Paraguay; Brazilian Amazons and northward); A. de Miranda
Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 287, fig. 113 (description, distribu¬
tion); Eigenmann, 1912, Mem. Carnegie Mus., 5: 175 (British Guiana: Creek below
Potaro Landing, Tumatumari, Rockstone, Crab Falls) ; A. de Miranda Ribeiro, 1914,
Comm. Linh. Telegr. Estr. Matto-Grosso ao Amazonas, 15, Annexo 5: 6 (Caceras, Rio
Paraguay, M. Grosso); Bertoni, 1914, Fauna Paraguaya: 7 (Alto Parana); Fisher, 1917,

132

ZOOLOGISCHE VERHANDELINGEN 132 (l 974 )

Ann. Carnegie Mus., 11: 412 (Asuncion, Rio Jauru) ; A. de Miranda Ribeiro, 1920,
Comm. Linh. Telegr. Estr. Matto-Grosso ao Amazonas, 58, annexo 5: 12 (Jamary?);
Beebe, 1925, Zoologica, 6: 117, 118 (Kartabo) ; Pearson, 1937, Proc. Calif. Acad. Sci.,
(4) 23: 92 (Balsas, Peru; Tingo de Pauca, Rio Maranon, Peru); Pearson, 1937, Proc.
Calif. Acad. Sci., (4) 23: ill (Paraguay basin); Fowler, 1941, Bol. Mus. “Javier Prado”,
5 : 375 (Peru (Rio Maranon)) ; Eigenmann & Allen, 1942, Fish. W. South America: 48,
53, 103 (Rio Paranapura, Yurimaguas) ; Schultz, 1944, Proc. U.S. Nat. Mus., 94: 203
(Venezuela); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 40 (Paraguai;
Amazonas e para o norte); Fowler, 1945, Los Peces del Peru: 41 (Peru (Rio Maranon)) ;
van der Stigchel, 1946, South American Nematognathi: 62 (Surinam; Lower Potaro
River, British Guiana); van der Stigchel, 1947, Zool. Meded., 27: 62 (Surinam; Lower
Potaro River, British Guiana); Fowler. 1051, Arq. Zool. S. Paulo, 6: 557 (Amazonia,
Peru, Bolivia, Rio Parana, Paraguay, Guiana, Venezuela) ; Boeseman, 1953, Zool. Meded.,
32: 7 (Marowini Basin, Nassau Mountains); Dahl, 1961, Noved. Colomb., 1: 494 (Co¬
lombia : the Guayabero River and several of its larger tributaries, such as Cano Morrocoy
and Cano Lozada) ; Luengo, 1963, Acta Biol. Venez., 3: 336 (Lago de Valencia, Vene¬
zuela, introduced); Lowe, 1964, J. Linn. Soc. Lond., Zool., 45: 110, 140, 143 (Rupununi
District); Mago-Leccia, 1967, Soc. Venez. Cienc. Nat. Bob, 27: 255 (Venezuela);
McConnell, 1967, Timehri, 43: 67 (Rupununi) ; Lundberg & Baskin, 1969, Amer. Mus.
Novit., 2398: 6, 29 (no locality).

Pseudorhamdia ornata ; Bleeker, 1864, Nat. Verh. Holl. Maatsch. Wetensch., (2) 20:
77 (Surinama).

Pimelodus ornata ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1:
168 (Goyaz); Eigenmann & Kennedy, 1903, Proc. Acad. Nat. Sci. Philad., 55: 499
(Paraguay: Asuncion, Arroyo Trementina); Eigenmann, McAtee & Ward, 1907, Ann.
Carnegie Mus., 4: 115 (Corumba, Matto Grosso).

Material. — 87 specimens as follows. One specimen, ca. 1824-1828, Cuyaba, Brazil (J.
Natterer, MV no. 45832), standard length 277 m, syntype of the species. One specimen,
1909, Suriname (Heller, MV no. 45841), standard length 195 mm. Five specimens, 22
December 1963, Tapoeripa Kreek near Brokopondo (Boeseman, RMNH nr. 25563),
standard length 139, 163, 203, 222, 230 mm. Four specimens, 3 January 1964, Affobakka,
below dam (Boeseman, RMNH no. 25564), standard length 230, 254, 256, 283 mm. Four
specimens, 4 January 1964, Suriname Rivier and creeks near Brokopondo (Boeseman,
RMNH no. 25565)., standard length 142, 145, 163, 170 mm. One specimen, 10 January
1964, near Brokopondo (Boeseman, RMNH no. 25566), standard length 130 mm. Four
specimens, 11 January 1964, creek near Brokopondo (Boeseman, RMNH no. 25567),
standard length 112, 122, 124, 138 mm. Four specimens, 12 January 1964, creek near
Brokopondo (Boeseman, RMNH no. 25568), standard length 114, 115, 138, 150 mm. One
specimen, 25 January 1964, Suriname Rivier near Brokopondo (Boeseman, RMNH
no. 25569), standard length 229 mm. Two specimens, 28 January 1964, creek near Broko¬
pondo (Boeseman, RMNH no. 25570), standard length 177, 245 mm. One specimen,
30 January 1964, Ceder Kreek near Brokopondo (Boeseman, RMNH no. 25571), standard
length 145 mm. One specimen 3/4 February 1964, Suriname Rivier near Brokopondo
(Boeseman, RMNH no. 25572), standard length 150 mm. Five specimens, 5 February
1964, Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25573), standard length
128, 128, 130, 156, 191 mm. Eight specimens, 6 February 1964, Suriname Rivier near
Brokopondo (Boeseman, RMNH no. 25574), standard length 114-194 mm. Two speci¬
mens, 7 February 1964, Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25575),
standard length 114, 192 mm. Seven specimens, same data (Boeseman, RMNH no. 25576),
standard length 144-256 mm. Eight specimens, 8 February 1964, Suriname Rivier near

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

133

Brokopondo (Boeseman, RMNH no. 25577), standard length 129-251 mm. One specimen,
14 February 1964, Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25578),
standard length 201 mm. Three specimens, 21 February 1964, Suriname Rivier near Bro¬
kopondo (Boeseman RMNH no. 25570), standard length 140, 149, 153 mm. Five speci¬
mens, 21 February 1964, Suriname Rivier near Brokopondo (Boeseman, RMNH
no. 25580), standard length 124, 126, 139, 141, 190 mm. One specimen, 27 February 1964,
Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25581), standard length
155 mm. One specimen, 20 March 1965, Suriname Rivier near Brokopondo (Mees, RMNH
no. 25582), standard length 154 mm. One specimen, same data (Mees, RMNH no. 26512),
standard length 173 mm. One specimen, 13 May 1965, Suriname Rivier near Botopassie
(Mees, RMNH no. 25670), standard length 61 mm. Five specimens, 13 August 1965,
Mamadam (Mees, RMNH no. 25583), standard length 113, 133, 177, 179, 254 mm. One
specimen, 2 September 1965, Suriname Rivier near Brokopondo (Mees, RMNH no. 25584),
standard length 207 mm. Five specimens, 23 January 1966, Sipaliwini (Mees, RMNH
no. 25585), standard length 156, 158, 170, 174, 242 mm. Two specimens, 6 February
1966, Sipaliwini (Mees, RMNH no. 25586), standard length 144, 157 mm. Two specimens,
9 February 1966, Sipaliwini (Mees, RMNH no. 25587), standard length 180, 225 mm.

Characters. — D 1 . 6 , A 11-14 (iii or iv.8J 4 to 10), P I.9 or 1 .10, V 6 (i.5),
C i.i5.i and rudiments, branchiostegals 10, gill-rakers 14-21, usually 15 to
18 (3 to 5 + 11 to 17). This species differs from P. blochii and P. albofascia -
tus by its more slender appearance, longer and anteriorly more depressed
head, wider mouth with upper jaw protruding much farther, different num¬
ber of gill-rakers, and most conspicuously by its colours as described below.

Colour. Upper surface of head, and back blackish-grey; a yellow band
above the lateral line, its anterior part curved upwards to the origin of the
dorsal fin; a broad blackish-grey band along the lateral line; remainder of
body pale, tinged yellowish; dorsal fin yellow, with a broad black median
band from the spine to the fourth or fifth ray; caudal fin yellow with on each
lobe a broad longitudinal blackish band; paired fins, anal and adipose fins
yellowish.

Whereas the yellow fades comparatively rapidly in preserved specimens
(which is undoubtedly the reason that it has not, to my knowledge, been
mentioned in literature), the black pigment remains, so that even specimens
collected over a century ago can readily be identified by it; especially the
black band or patch on the dorsal fin is an easy character.

Habitat. — A common inhabitant of greater and smaller rivers, not often
found in creeks.

Distribution. — This species is widely distributed in tropical South America
from Colombia (Guayabero basin) and Venezuela (Calabozo, Orinoco basin),
through the Guianas and the whole Amazon basin to the Rio Paraguay.

Discussion. — Its characteristic colour pattern, well described and illus¬
trated by Kner (1858), causes this to be an easily-recognized species, which
has for this reason remained largely clear of synonyms and misidentifications.

*34

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

The one synonym, Megalonema rhabdostigma, has long since been recognized
as such (cf. Schultz, 1944a: 215).

Pimelodus blochii Valenciennes
(Pimelodus clarias Auct.)

Pimelodus Blochii Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.), 15: 139 — Cayenne, Surinam, Colombie; here restricted to Suriname.

Pseudorhamdia macronema Bleeker, 1864, Nat. Verh. Holl. Maatsch. Wetensch., (2)
20: 79, pi. 14 (Surinama).

Silurus Clarias ; (pt.) Bloch, 1785, Ichthyologie, 1: 198, pi. 35 fig. 1 (du Bresil & de
Surinam).

Silurus Callarias ; (pt.) Bloch & Schneider, 1801, Syst. Ichth.: 379 (in fluviis Americae
australis).

Pimelodus clarias ; (pt.) La Cepede, 1803, Hist. Nat. Poiss., 5: 93, 100 (eaux douces
du Bresil et dans celles de Surinam).

Pimelodus clarias ; (pt.) van der Stigchel, 1946, South American Nematognathi: 59
(Surinam, Paramaribo); (pt.) van der Stigchel, 1947, Zool. Meded., 27: 59 (Surinam,
Paramaribo) ; Boeseman, 1953, Zool. Meded., 32: 7 (Sommelsdijkse Kreek, near Wolfen-
buttel (asylum), Paramaribo; Paramaribo, Surinam River; Surinam River, near planta¬
tion Peperpot; Surinam River, near Jodensavanne).

Piramutana Blochii ; Kappler, 1881, Hollandisch-Guiana: 167 (Hollandisch-Guiana);
Kappler, 1885, Das Ausland, 58: 919 (Hollandisch Guiana) ; Kappler, 1887, Surinam: 152
(Surinam).

Pimelodes clarias ; Lijding, 1959, Surinaamse Landb., 7: 73 (Suriname).

Material. — One specimen, 1824-1836, Suriname, presumably Suriname Rivier near
Paramaribo (H. H. Dieperink, RMNH no. 3069), standard length 177 mm, figured syn-
type of Pseudorhamdia macronema Bleeker. One specimen, same data (H. H. Dieperink,
RMNH no. 3070), standard length 128 mm, syntype of Pseudorhamdia macronema . One
specimen, same data (H. H. Dieperink, RMNH no. 3071), standard length 147 mm,
syntype of Pseudorhamdia macronema. One specimen, same data (H. H. Dieperink,
RMNH no. 3072), standard length 97 mm, syntype of Pseudorhamdia macronema. One
specimen, same data (H. H. Dieperink, MNP no. 72), standard length 216 mm, very
likely syntype of Pimelodus Blochii. Two specimens, 5 January 1964, Suriname Rivier
and creeks near Brokopondo (Boeseman, RMNH no. 26503), standard length 103,
139 mm. One specimen, 12 January 1964, tributary of Suriname Rivier near Brokopondo
(Boeseman, RMNH no. 26504), standard length 97 mm. Three specimens, 3/4 February
1964, Suriname Rivier near Brokopondo (Boeseman, RMNH no. 26499), standard length
170, 175, 195 mm. 20 specimens, 5 February 1964, Suriname Rivier near Brokopondo
(Boeseman, RMNH no. 26500), standard length 84-168 mm. One specimen, 7 February
1964, Suriname Rivier near Brokopondo (Boeseman, RMNH no. 26502), standard length
154 mm. Two specimens, 8 February 1964, Suriname Rivier near Brokopondo (Boese¬
man, RMNH no. 26501), standard length 172, 172 mm. One specimen, 14 February 1964,
Suriname Rivier near Brokopondo (Boeseman, RMNH no. 26507), standard length
125 mm. Two specimens, 21 February 1964, Suriname Rivier near Brokopondo (Boese¬
man, RMNH no. 26506), standard length 104, 134 mm. Three specimens, 27 February
1964, Suriname Rivier near Brokopondo (Boeseman, RMNH no. 26505), standard length
I 55 » 156, 165 mm. One specimen, same data (Boeseman, RMNH no. 26508), standard
length 128 mm. One specimen, 29 May 1964, Suriname Rivier near Brokopondo (Boese¬
man, RMNH no. 26510), standard length 89 mm. One specimen, 3 June 1964, Suriname
Rivier near Brokopondo (Boeseman, RMNH no. 26509), standard length 76 mm. Four
specimens, 20 March 1965, Suriname Rivier near Brokopondo (Mees, RMNH no. 26513),

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

135

standard length 104, 115, 116, 178 mm. One specimen, 18 April 1967, Lokka-Lokka
tabbetje, Marowijne, 6 km S. of mouth Gran Kreek (Nijssen, ZMA no. 106837), standard
length 113 mm. Four specimens, 23 June 1966, mouth of Coppename and Saramacca
(Nijssen, ZMA no. 106838), standard length 94, 116, 119, 168 mm. Two specimens, Sara¬
macca Rivier near Coppename Punt, 23 Juni 1966 (Nijssen, ZMA no. 106840), standard
length 84, 88 mm. Five specimens, 23 June 1966, mouth of Coppename and Saramacca
(W. Vervoort, RMNH no. 26511), standard length 92, 129, 135, 137, 181 mm.

Characters. — D 1 . 6 , A 12-13 (usually iv.8, also v.8, iv.9, and twice iv.io),
P I. 9, occasionally I.io 1 ), V 6 (i.5), C i.i5-i x ) and rudiments, branchio-
stegals 8, gill-rakers 24-27, once 30 (5 to 7, once 9, + 1 + 18 to 20). Diameter
of eye 1.0-1.7 times in bony interorbital 2 ).

Colour. Dorsal surface uniform light liver-brown, under parts white, un-
pigmented, the flanks usually with a silvery gloss, a narrow white streak along
the lateral line, base of dorsal spine blackish. See Bleeker’s (1864) excellent
coloured plate.

Distribution and habitat. — A species of the lower courses of the great
rivers and the river mouths, where it is evidently common. The species occurs
probably in all large rivers of Suriname, but is at present known only from
the Marowijne, the lower course of the Suriname Rivier and the mouth of
the Saramacca-Coppename. One would expect a species with this kind of
habitat to have a wide distribution in adjacent parts of South America, and
such is probably the case, but because of insufficiently known geographical
variation, published records cannot be accepted without a renewed study of
the material they are based on. Eigenmann’s (1912: 174) variety A from
coastal British Guiana is almost certainly this species; this wa's also Eigen¬
mann’s own opinion, for he wrote under variety A: “Here belongs also the
macronema of Bleeker”. The situation in British Guiana is, however,
extremely complicated, as discussed under the following species.

Discussion. — Valenciennes (in Cuvier & Valenciennes, 1840: 125-128,
139-140) has stated clearly why the specific name clarias is not applicable
to this species, and it is surprising that its use has been continued right up to
the present time. The name Pimelodus maculatus has also been used for
fishes from Suriname, but the type-locality is the Rio de la Plata, and indivi¬
duals from there differ from Suriname fishes, if by nothing else, at least by
having black blotches on the sides, as clearly described by La Cepede (1803:

1) One specimen (RMNH no. 26506, 134 mm) has the right pectoral ray 1.9, the left
one I.7; one specimen (RMNH no. 265x1, 92 mm) has C i.i4.i.

2) The longest, horizontal diameter of the externally visible eye was measured, hence
from the soft rim, and this taken in the bony interorbital. Only two specimens in the
series measured have the eye as large as 1.0 (a specimen of 156 mm standard length,
RMNH no. 26505, and one of 84 mm standard length, RMNH no. 26500).

Unexpectedly, there is no clear relation between body-size and relative size of the eye.

136

ZOOLOGISCHE VERHANDELINGEN 13 2 (1974)

94, 107), and also present in a specimen from La Plata in Leiden (van der
Stigchel, 1946: 62). The two La Plata specimens in Leiden (RMNH no.
17302) also differ by their small and narrow dorsal plate; the eyes are small,
but variable: the spotted fish has them 1.4 times in interorbital, the other one,
which has much smaller eyes as well as a broader head, 2.0. Perhaps a sexual
difference, as the small-eyed immaculate specimen has also a slightly heavier
body, but possibly specific or subspecific differences are involved (cf. Devin-
cenzi & Teague, 1942: 44-45).

The first name definitely applicable to the Suriname species is Pimelodus
blochii ; this name was partly based on Bloch (1785, pi. 35), partly on material
personally examined by Valenciennes. Bloch mentioned material from Brazil
and Suriname, Valenciennes wrote: “Ce poisson ... est de l’Amerique
equinoxiale, et particulierement de la Guiane. Nos echantillons viennent tous
ou de Cayenne, ou de Surinam, et nous en avons vu dans les mains deM.de
docteur Roulin des figures faites en Colombie”. In order to eliminate a possible
source of confusion, I restrict the type-locality of Pimelodus blochii to Suri¬
name. It is likely that the locality Suriname was based on material collected
by Dieperink, and as Dieperink lived in Paramaribo, the type-locality can
further be restricted to the lower course of the Suriname Rivier. Therefore
I treat material from the Suriname Rivier as topotypical, and base upon it
comparisons with material from other river systems.

I do not think that in restricting the type-locality of P . blochii , I need be
concerned about Schultz (1944a: 206), who stated that he selected the Mag¬
dalena River, Colombia, as representing the type-locality of “Pimelodus
clarias clarias (Bloch)”. Apart from the fact that the name has no status, as
it was only a misapplication of Silurus clarias Hasselquist, Colombia was not
included in the range of the species as given by Bloch. In the discussion of the
following species, evidence will be presented that the population inhabiting
the Rio Magdalena is not identical with that of Suriname.

Boeseman’s (1953: 7) remark that his specimens, which I re-examined:
“In several characters ... seem to be intermediate between Bloch’s clarias and
the subspecies coprophagus Schultz”, is puzzling inasmuch as these specimens
are topotypical of “Bloch’s clarias ”, and the material from the Suriname Rivier
is homogeneous in appearance. Dr. Boeseman (pers. comm.) meant to express
doubt of the validity of the characters given by Schultz, and in this he was
certainly justified. It is not impossible that Pimelodus rigidus Spix (1829:
19, pi. VII fig. 2) from “Brasilia aequatoriali” is the same as P. blochii , but
in view of the apparently pronounced geographical variation occurring in this
group of forms, I believe that this question could only be solved by a study
of fresh topotypical material.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

137

Note. — In the mouth-cavity of one specimen (RMNH no. 26511, 92 mm),
a crustacean, Telotha henselii (von Martens), of the family Cymathoidae was
found (RMNH no. Isop. 1703), provisionally identified by Dr. L. B. Holt-
huis.

Pimelodus albofasciatus species nova (pi. 4)

Material. — One specimen, 6 March 1961, Sipaliwini (H. P. Pijpers, ZMA no. 106839),
standard length 122 mm. Two specimens, 2 February 1966, Sipaliwini (Mees, RMNH
no. 26520), standard length no, 112 mm. One specimen, 4 February 1966, Sipaliwini
(Mees RMNH no. 26516), standard length 128 mm, holotype of the species. Seven
specimens, same data (Mees, RMNH no. 26519), standard length 99-114 mm. Ten speci¬
mens, same data (Mees, RMNH no. 26518), standard length 103-121 mm. 35 specimens,
6 February 1966, Sipaliwini (Mees, RMNH no. 26514), standard length 92-137 mm.
13 specimens, 9 February 1966, Sipaliwini (Mees, RMNH no. 26517), standard length
96-131 mm. One specimen, 21 June 1968, Coeroeni (M. S. Hoogmoed, RMNH no. 26515),
standard length 123 mm.

Characters. — D 1 . 6 , A 11-13 (iii to v.8), P I.9, less often 1 .10 (once 1.8
on one side), V 6 (i.5), C i.15.1 and rudiments, branchiostegals 8, some¬
times 9 (once 7 on one side, 8 on the other), gill-rakers 21-26, usually 23 or
24 (4 to 6 + x + 15 to 19). Morphologically extremely dose to P. blochii,
but eye larger: diameter of eye greater than width of bony interoibital,
o.7-1.0 times in its width; the number of gill-rakers averages one or two
lower; serrations along anterior edge of pectoral spine usually weaker, to
almost absent (this is an average difference only). Moreover there is a
highly characteristic colour pattern of dark and pale bands, as described in
the following paragraph.

Colour. Upper parts dark grey, with some slightly paler, indistinct, wavy
longitudinal lines; a broad white longitudinal band from head to tail, which
includes the lateral line over its whole length; below this white band is a
slightly broader dark grey band of equal length; a short grey band from the
posterior edge of the scapula, narrowing backwards and ending above and
just behind the ventral base; base of dorsal spine blackish; all fins slightly
pigmented, dusky, the dorsal spine more strongly pigmented, and the adipose
fin usually with some roundish grey dots; dorsal surface of maxillary barbels
grey.

Distribution. — Known from the upper course of the Sipaliwini, where
the species was common during my stay in February 1966, from the Coeroeni
near the aerodrome, where a single specimen has been obtained, and from
the upper Corantijn (or New River).

Discussion. — Apart from the slight differences in morphology and colour
mentioned above, in Suriname there is also a distinct ecological difference
between the species P. blochii and P. albofasciatus . The former occurs in the

138 ZOOLOGISCHE VERHANDELINGEN I32 (1974)

lower courses and the mouths of the great rivers: there are many records
from the coastal area, but in the now reasonably well-known Suriname Rivier
we found it inland only to Brokopondo. Although in some of the large rivers
it may go farther upstream, it appears definitely to keep to large bodies of
water. P. albofasciatus, on the other hand, occurs in a river where the largest
pools were rarely more then ten metres wide, after a few days of dry weather
one could jump over the Sipaliwini in several places. The Coeroeni specimen
was not taken in the river but in a shallow little pool near it. To what extent
this ecological difference exists also in adjacent countries remains to be seen,
but it looks as if in Essequibo the situation is more complicated than in Suri¬
name. Eigenmann (1912: 172-175) already distinguished several ‘Varieties”.
Specimens have been recorded from the Rupununi under the name P. clarias
by Fowler (1914: 263) and Lowe (1964: no, in, 113); these records were
not accompanied by descriptions, but Mrs. McConnell has sent me her mate¬
rial from British Guiana: two specimens, September 1957, Manari Creek
(Amazon drainage), standard length 66, 90 mm. — One specimen, 9 Decem¬
ber 1957, Dadanawa, Rupununi South Savanna, standard length 156 mm. —
One specimen, 1957, Rupununi, standard length 134 mm. — Two specimens,
22 April 1961, sandcreek road crossing, Rupununi River, standard length
87, 108 mm. These fishes have the same colour pattern as P. albofasciatus,
to which they are extremely similar in all characters except that the eyes are
smaller, 0.8-1.65 times in interorbital, and the spines tend to be longer (dorsal
spine equal to head or longer, against usually equal to head of shorter in
P. albofasciatus). Eigenmann’s (1912: 175) two Rupununi specimens, which
he described as “especially bright” were evidently similar.

Different from anything I have seen from Suriname is a specimen collected
on 9 July 1958, Lama conservancy, Lamaha Canal behind Georgetown (R. H.
McConnell), standard length 176 mm. This fish is spotted all over, with small
spots on the head; larger spots in five irregular rows above the lateral line,
and irregular and less distinct ones below it; lateral line narrow white. This
corresponds with Eigenmann’s description of specimens from Lama Stop-
Off and the Botanic Gardens. The colour-photograph published by Hoede-
man (1954) shows a fish of similar appearance. Hoedeman wrote that this
fish came from Suriname, and he may be right, but it appears preferable
not to admit this “variety” to the ichthyofauna of Suriname on the basis of
an aquarium-fish.

Two specimens from Bartica, Essequibo River (C. H. Eigenmann, BM
no. 1911.10.31: 44-48), standard length 81, 97 mm, are very similar to P.
blochii, but above the lateral line there are some vague and wavy vertical pale
lines, not quite sufficiently pronounced to give the fishes a marbled appear-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

139

ance. Mrs. McConnell suggested that these might be intermediate between
P. blochii and the Rupununi specimens. This, however, is doubtful: the dorsal
spines are shorter than the head, strongly serrated along their anterior border
(much stronger than in the Rupununi specimens), and the colour-pattern just
described cannot be called intermediate either. On the other hand Eigen-
mann’s (1912: pi. 74) map does suggest that in the Essequibo basin the
distribution of Pimelodus “clarias” is continuous from the coast to the Rupu¬
nuni, and intermediate populations may well occur. Compare the situation
as it is now known to exist in Suriname and may exist in Essequibo, with
that described by Haseman (1911: 380) for the Rio Iguassu, where Pimelo¬
dus ortmanni occurs in the headwaters, Pimelodus “clarias” in the lower
course.

Its description indicated that P. clarias coprophagus Schultz from western
Venezuela has a colour-pattern not unlike P. albofasciatus, and for that
reason a direct comparison was desirable. Two paratypes of this form were
available to me for this purpose, standard length 135, 169 mm (L. P. Schultz,
6 March 1942, Lago Maracaibo, ZMA no. 102124, ex USNM no. 121154).
Their colour-pattern is indeed fairly close to that of P. albofasciatus, but it is
less regular, the white band along the lateral line is neither so wide nor so
regular, and the pigmented areas tend to break up in blotches and dots. Also
the eyes are remarkably smaller, 1.6 and 2.0 times in interorbital, and there
are more gill-rakers. Ecologically P. c. cophrophagus differs from P. albo¬
fasciatus in that it is a coastal form; in this it agrees with P. blochii.

The number of gill-rakers deserves further discussion. In the specimens
of P. c. coprophagus it is 8 + 1 +22 and 7 + 1 + 22, total number on the
outer branchial arch 31 and 30. This tallies with the figures given by Schultz
(1944a, table 2) for a much larger number of individuals. The number agrees
also with the counts given by Schultz for specimens ascribed by him to P.
clarias clarias (= P. blochii ), from the Rio Magdalena, Colombia. Topo-
typical P. blochii from Suriname, on the other hand, normally has 24-26
gill-rakers, although in one specimen I counted as many as 30. In P. albo¬
fasciatus also, 24-25 is the commonest number. From this I deduce that
Schultz’s specimens from the Rio Magdalena are not identical with topo-
typical P. blochii . Lacking material from the Rio Magdalena, I have to leave
the matter at this.

Eigenmann (1912: 176, pi. XVI fig. 2) described Pimelodus heteropleurus
from Rupununi Pan as a: “Form of Pimelodus clarias ”. The type specimen,
of only 46 mm standard length, remains unique, but judging by its descrip¬
tion and plate, it differs clearly from Mrs. McConnell’s Rupununi specimens
of Pimelodus cf. albofasciatus. As the latter is evidently the local represen-

140

ZOOLOGISCHE VERHANDELINGEN 1 32 (1974)

tative of the P. maculatus ( — P. clarias auct.) group, P. heteropleurus cannot
belong to the same group.

Pimelodus sp.

Material. — One specimen, 21 March 1968, Xaventina, Rio das Mortes, Mato Grosso
(R. H. McConnell, BM no. 19717.29: 72), standard length 147 mm. One specimen,
23 March 1968, same locality (R. H. McConnell, BM no. 1971.7.29: 73), standard length
121 mm. Two specimens, 12 May 1968, Suia Missu, Mato Grosso (R. H. McConnell, one
BM no. 1971.7.29: 74, the other RMNH no. 26601), standard length 142, 170 mm.

Characters. — D 1 . 6 , A 11-12 (iii or iv.8 or 9), P I.9, V 6 (i.5), C i.is.i
and rudiments, gill-rakers 20-21 (4 or 5 + 1 + 15). Predorsal length 2.6
times in standard length; eye 3.8-4.0 times in head, 0.95-1.0 in interorbital;
maxillary barbels reaching at least to base of caudal fin or to beyond tip of
tail; outer pair of mandibular barbels variable in length, from not reaching tip
of pectorals to almost reaching tip of ventrals; spines long, dorsal spine
0.8-0.9 in head; anterior edge of pectoral with moderate serrations, its distal
third smooth; base of adipose fin 4.75-5.5 in standard length. Morphologically
this form is very close to the two preceding species, but the number of gill-
rakers is distinctly lower, the spines are longer (dorsal spine longer than
head, against equal to head or shorter in the other species), and the colour
and colour-pattern are different, as described in the next paragraph.

Colour. Dark bronze-brown; under surface white, unpigmented; a narrow
but very distinct white streak along the lateral line; the smallest specimen
but none of the others has a pale brown streak between the lateral line and
the back, following the dorsal profile; a black spot at the base of the dorsal
spine as in the other species; fins only lightly pigmented, except for the spine
and first ray of the dorsal fin which are dark.

Distribution. — Known from the Rio das Mortes (a tributary of the Ara-
guaia) and the Suia Missu (a tributary of the Xingu), Mato Grosso, Brazil.
It is interesting that these rivers appear to have the same form.

Discussion. — Only two of the fishes have spines; in the other two, from
Xaventina, dorsal as well as pectoral spines have been chopped off near their
bases. These specimens have evidently been collected by Indians, and they
also show cutlass-damage. One had been cut across the head in such a way
that the orbit could not be measured.

It may appear inconsistent that I have named P. albofasciatus, but refrain
from giving this form, which is as well differentiated from other known
members of this group, a name. The reason is that whereas in Suriname the
position is now reasonably clear, with a coastal and an interior form, the
naming of a single form in this complex group from the middle of Amazo-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE I4I

nia would in my opinion be premature before material collected over a wide
area becomes available.

Pimelodus altissimus Eigenmann & Pearson
(Pimelodus altipinnis Auct., nec Steindachner, 1864)

Pimelodus altissinus Eigenmann & Pearson, 1942, in Eigenmann & Allen, Fish. W.
South America: 106, pi. V fig. 5 — Rio Ucayali, near Orellana.

Pimelodus eigenmanni van der Stigchel, 1946, South American Nematognathi: 64 —
Brazil.

Pimelodus altipinnis ; Steindachner, 1876, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw.
Cl., 72 (x): 60S, pl- XI (Amazonenstrom hei Para, Santarem und Cameta) ; Eigenmann &
Eigenmann, 1888, Proc. Calif. Acad. Sci., (2) 1: 135 (Para) ; Eigenmann & Eigenmann,
1890, Occ. Pap. Calif. Acad. Sci., 1: 180 (Para) ; (pt.) Eigenmann & Eigenmann, 1891,
Proc. U.S. Nat. Mus., 14: 30 (Amazon); Perugia, 1897, Ann. Mus. Genova, 38: 18
(Alto beni, Bolivia); Goeldi, 1898, Bol. Mus. Paraense, 2: 461, 476 (Magoary); Eigen¬
mann & Bean, 1907, Proc. U.S. Nat. Mus., 31: 660 (Amazon River between Para and
Manaos); Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 388 (Amazon
and northward); A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16:
293, fig. 114 (Amazonas); Starks, 1913, Stanford Publ. Univ. Ser.: 27 (Para); A. de
Miranda Ribeiro, 1914, Comm. Linh. Telegr. Estr. Matto-Grosso ao Amazonas, 15,
Annexo 5: 6 (Manaos); Fisher, 1917, Ann. Carnegie Mus., 11: 415 (Belem, Para, San
Joaquin, Santarem); Pearson, 1924, Indiana Univ. Stud., 11 (64): 7 (Beni basin);
Pearson, 1937, Proc. Calif. Acad. Sci., (4) 23: in (Mamore basin) ; Fowler, 1940, Proc.
Acad. Nat. Sci. Philad., 92: 95 (Bolivia: R. Beni; San Joaquin); Delsman, 1941, Verh.
Kon. Natuurhist. Mus. Belgie, (2) 21 (3) : 79 (Trombetas-river, near Obidos) ; Gosline,
1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 41 (Amazonas e para o norte) ; Fowler,
1951, Arq. Zool. S. Paulo, 6: 553, fig. 568 (Amazonia, Peru, Bolivia, Guiana).

Pimelodus altissimus; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 42 (rio
Ucayali); Fowler, 1945, Los Peces del Peru: 41 (Peru (Cerca de Orellana)); Fowler,
1951, Arq. Zool. S. Paulo 6: 554 (Alto Amazonas, Peru).

Material. — One specimen, 1893, Rio Beni, Bolivia (L. Balzan, MCG no. 8122),
standard length 194 mm. One specimen, no date, Brazil (RMNH no. 15465, ex Mus.
Rio de Janeiro), standard length 140 mm, virtual holotype of Pimelodus eigenmanni.

Discussion. — Van der Stigchel (1946: 64) described Pimelodus eigen¬
manni “nom. nov.”, ending his description with the words: "... and therefore
I have introduced the new name Pimelodus eigenmanni”. To the casual reader
this obviously suggests that a new name was introduced, to replace one pre¬
viously given to the same species, that was for some reason unavailable. Closer
scrutiny reveals however, that Pimelodus eigenmanni is a new species, based
on a specimen in Leiden, with which van der Stigchel subjectively identified
a number of references in literature to Pimelodus altipinnis Steindachner.
Van der Stigchel was of the opinion that previous authors had misidentified
specimens as P. altipinnis , which actually belonged to his new species. Van
der Stigchel’s specimen is indeed a Pimelodus, and it differs clearly from

142

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Pimelodus altipinnis which, as I shall indicate on a later page, is a synonym
of Pimelodella cristata.

Van der Stigchel (1946) probably took the incorrect procedure of indicating
his new species by the name “nom. nov.” from Eigenmann (1917: 237), who
named Pimelodella steindachneri in a similar way.

Pimelodus eigenmanni van der Stigchel is a primary homonym of Pime¬
lodus (Pimelodella) eigenmanni Boulenger (1891: 232, footnote 3), and
therefore is not an available name for this species. I do not consider that a
new name is required, as Pimelodus altissimus Eigenmann & Pearson is
obviously the same species. Admittedly its describers differentiated P. altissi¬
mus as being: “Closely related to P. altipinnis, from which it differs by pos¬
sessing a much higher dorsal spine, longer pectoral spines, and a steeper pre¬
dorsal profile”, but it is likely that these rather insignificant differences are
within the limits of individual variation. P. altissimus was based on a single
specimen, and it is not clear whether a comparison was made with actual speci¬
mens of “P. altipinnis ”, or only descriptions and Steindachner’s (1876a: pi.
XI) plate were consulted.

Genus Pimelodella Eigenmann & Eigenmann

Pimelodella Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2) 1: 131 — type
by original designation, Pimelodus cristatus Muller & Troschel.

Generic diagnosis. — The main character on which this genus was founded
is a narrow occipital process, of uniform width throughout, meeting the
dorsal plate (fig. 33). In addition the species are slender, with a not very
large head, the bones of which are covered with skin; long maxillary bar¬
bels; well-developed eyes with free rim; teeth in a band in each jaw, without
backward projections; pungent pectoral spines; well-developed, usually very
long, adipose fin, and forked tail.

Discussion. — Pimelodella appears to be one of the largest genera known;
Eigenmann (1917) recognized 34 species, Gosline (1945) listed no fewer
than 54 species, and more have been described since. Eigenmann (1912)
recorded three species from British Guiana. In this connection it is sur¬
prising that in Suriname we have been unable to find more than one, P,
cristata, which belongs to the commonest and most widely distributed species
of the country. A second species, C. macturki is known in Suriname from a
single specimen, without exact locality. Kappler’s (1885, 1887) record of
“Pimelodus lateristriga from Suriname is evidently based on a misidentifi-
cation of P. cristata , which he did not mention. P . lateristriga is not known
to occur in the Guianas.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

143

Fig. 33. Heads viewed from above to show occipital process and extent of fontanel.
Pimelodella cristata (top left), Imparfinis minutus (bottom left), Rhamdia quelen (top

right) and Rhamdia foina (bottom right). Imparfinis 2 X, the others natural size.

A review of the many described species is out of place here, but I would
like to draw attention to the fact that in some instances pairs of species have
been described from one river, differing mainly in size of the eye (e.g., Fowler,
1940b: 75-77; 1941a: 127-135). The discovery of a sexual difference in eye-
size in P. cristata, makes me suspect that at least some of such pairs are based
on the two sexes of one species. The only sexual difference previously de¬
scribed in the genus is in the dorsal fin which in the males of some species
has a filament, not found in females and juveniles (cf. Dahl, 1961: 498; van
der Stigchel, 1964).

144

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Pimelodella cristata (Muller & Troschel) (fig. 33)

P[imelodus] cristatus Muller & Troschel, 1848, in Schomburgk, Reisen in Britisch-
Guiana, 3: 628 — Takutu und Mahu.

Pimelodus altipinnis Steindachner, 1864, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw.
Cl., 49 (1) : 213, pi. II fig. 3, 4 — Demerara.

Pimelodus (Pseudorhamdia) IVesselii Steindachner, 1876, Sitzb. Akad. Wiss. Wien,
Mathem.-Naturw. Cl., 74 (1) : 614, footnote 1 — Essequibo.

Pimelodella geryi Hoedeman, 1961, Bull. Aquatic Biol., 2: 134 — French Guiana:
mainland, Litany river, village Aloike = Alowike, Suriname.

Pimelodus crislatus ; Muller & Troschel, 1849, Horae Ichth., 3: 4 (Guiana im Esse¬
quibo); Gunther, 1864, Cat. Fish. Brit. Mus., 5: 117 (Guiana; River Capin, Para; Esse¬
quibo River); Puyo, 1949, Poiss. Guyane Fr.: 88, fig. 49 (les “criques” ... du cours
superieur de l’Itany).

Pimelodus lateristriga; Kappler, 1885, Das Ausland, 58: 919 (Hollandisch Guiana);
Kappler, 1887, Surinam: 152 (Surinam).

Pimelodus gracilis ; Vaillant, 1899, Bull. Mus. Hist. Nat. Paris, 5: 155 (le haut du
fleuve Carcevenne: riviere Lunier, riviere Carnot); Vaillant, 1900, Nouv. Arch. Mus.
Paris, (4) 2: 124, 127 (riviere Carnot).

Pimelodella cristata ; Eigenmann, 1912, Mem. Carnegie Mus., 5: 168 (British Guiana:
Tumatumari, Below Packeoo Falls, Rockstone, Konawaruk, Creek below Potaro Landing,
Twoca Pan); Fowler, 1914, Proc. Acad. Nat. Sci. Philad., 66: 263 (Rupununi River);
Eigenmann, 1917, Mem. Carnegie Mus., 7: 236 (synonymy and distribution) ; Beebe,
1925, Zoologica, 6: 118, 120 (Kartabo); van der Stigchel, 1946, South American Nemato-
gnathi: 54 (upstream from Gran Rio; Pikien River, Tresling; Potaro Landing, Lower
Potaro River) ; van der Stigchel, 1947, Zool. Meded., 27: 54 (upstream from Gran Rio;
Pikien River, Tresling; Potaro Landing, Lower Potaro River); Boeseman, 1953, Zool.
Meded, 32: 2, 6 (Moengo Tapoe, Djaikreek).

Pimelodus altipinnis ; Eigenmann, 1912, Mem. Carnegie Mus., 5: 177 (Demerara).
Pimelodella gracife ; Fowler, 1914, Proc. Acad. Nat. Sci. Philad., 66: 263, fig. 13
(Rupununi River).

Pimelodella sp.; Lowe, 1964, J. Linn. Soc. Lond., Zool., 45: no, 115, 140, 143 (Rupu¬
nuni District).

Material. — 974 specimens from Suriname as follows. One specimen, 27 November
1957, Alowike, Litani, Suriname (J. Gery, ZMA no. 102235), standard length 59 mm,
holotype of Pimelodella geryi. Three specimens, 12 December 1963, creek near Drepada,
ca. 8 km N. of Brokopondo (Boeseman, RMNH no. 25672), standard length 83, 89,
91 mm. One specimen, 26 December 1963, Sarakreek near Dam (Boeseman, RMNH
no. 25673), standard length 85 mm. 70 specimens, 28 December 1963, Kwambaolo Kreek
near Dam (Boeseman, RMNH no. 25674), standard length 69-176 mm. 53 specimens, same
data (Boeseman, RMNH no. 25675), standard length 68-136 mm. 51 specimens
same data (Boeseman, RMNH no. 25676), standard length 77-173 mm. One specimen,
25 December 1963, Sara Kreek between Locus and Dam (Boeseman, RMNH no. 25677),
standard length 108 mm. One specimen, 4 January 1964, Suriname Rivier near Broko¬
pondo (Boeseman, RMNH no. 25678), standard length 103 mm. One specimen, 5 January
1964, creek near Brokopondo (Boeseman, RMNH no. 25679), standard length 92 mm.
Three specimens, 7 January 1964, Suriname Rivier, rapids at Brokopondo (Boeseman,
RMNH no. 25680), standard length 105, 106, 112 mm. One specimen, 9 January 1964,
Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25681), standard length
88 mm. Four specimens, 10 January 1964, near Brokopondo (Boeseman, RMNH
no. 25682), standard length 80, 83, 87, 95 mm. Three specimens, 11 January 1964, tribu¬
tary of Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25683), standard

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

145

length 91, 102, 114 mm. Three specimens, 12 January 1964, tributary of Suriname Rivier
near Brokopondo (Boeseman, RMNH no. 25684), standard length 80, 96, 142 mm. Three
specimens, 16-17 January 1964, Mamadam (rapids), Suriname Rivier (Boeseman, RMNH
no. 25685), standard length 109, no, 113 mm. Four specimens, 16-17 January 1964,
Suriname Rivier below Mamadam (Boeseman, RMNH no. 25686), standard length 91,
96, 102, 140 mm. Three specimens, 28 January 1964, creek near Brokopondo (Boeseman,
RMNH no. 25687), standard length 93, 112, 116 mm. Four specimens, 29 January 1964,
Ceder Kreek near Brokopondo (Boeseman, RMNH no. 25688), standard length 114,
114, 115, 117 mm. Two specimens, 30 January 1964, Ceder Kreek near Brokopondo (Boe¬
seman, RMNH no. 25689), standard length hi, 115 mm. Seven specimens, 3-4 February
1964, Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25690), standard length
82-132 mm. 27 specimens, 5 February 1964, Suriname Rivier near Brokopondo (Boese¬
man, RMNH no. 25691), standard length 62^-120 mm. 43 specimens, 6 February 1964,
Suriname Rivier downstream from Brokopondo (Boeseman, RMNH no. 25692), standard
length 67-139 mm. 27 specimens, 7 February 1964, Suriname Rivier near Brokopondo
(Boeseman, RMNH no. 25693), standard length 76-139 mm. 30 specimens, 8 February
1964. Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25694), standard length
79-136 mm. Ten specimens, 12 February 1964, rapids Suriname Rivier near Brokopondo
(Boeseman, RMNH no. 25695), standard length 60-122 mm. Two specimens, 14 February
1964, Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25696), standard length
85, 86^2 mm. Seven specimens, 18 February 1964, Jabocai Kreek and Witte Kreek,
tributaries of the Suriname Rivier between Kabel und Lombe (Boeseman, RMNH
no. 25697), standard length 75-92 mm. 16 specimens, 21 February 1964, Suriname Rivier
near Brokopondo (Boeseman, RMNH no. 25698), standard length 73-125 mm. Five
specimens, 27 February 1964, Suriname Rivier near Brokopondo (Boeseman, RMNH
no. 25699), standard length 85-113 mm. Three specimens, 5 March 1964, Gran Kreek
near Bofroedede (Boeseman, RMNH no. 25700), standard length 87, 107, 119 mm.
One specimen, 6 March 1964, Gran Kreek near its mouth (Boeseman, RMNH no. 25701),
standard length 82 mm. 21 specimens, 6 March 1964, rapids in Gran Kreek near mouth
(Boeseman, RMNH no. 25702), standard length 66-120 mm. 16 specimens, 9 April 1964,
pools downstream from Feddiprati rapids, Saramacca Rivier (Boeseman, RMNH
no. 25703), standard length 69-104 mm. 11 specimens, 12 May 1964, Suriname Rivier near
Brokopondo (Boeseman, RMNH no. 25704), standard length 87-118 mm. Four specimens,
3 June 1964, Suriname Rivier near Brokopondo (Boeseman, RMNH no. 25705), standard
length 75, 77, 78, 07 mm. One specimen, 5 June 1964, Brokopondo Meer near mouth Sara
Kreek (Boeseman, RMNH no. 25706), standard length 102 mm. One specimen, 23 June

1964, Suriname Rivier near Aloesoebanjivallen (Boeseman, RMNH no. 25707), standard
length 37 mm. 35 specimens, 21 July 1964, tributary of the Gran Kreek, ca. 12 km above
its mouth (Boeseman, RMNH no. 25708), standard length 25-46 mm. One specimen,
same data (Boeseman, RMNH no. 25709), standard length 151 mm. Eight specimens,
29 July 1964, pool at middle course Gran Kreek (Boeseman, RMNH no. 25710), standard
length 81-89 mm - 95 specimens, 30 July 1964, Gran Kreek, near future shore of lake
(Boeseman, RMNH no. 25711), standard length 25-57 (93 ex.), 118, 118 mm. 13 speci¬
mens, 31 July 1964, rapids in middle course Gran Kreek (Boeseman, RMNH no. 25712),
standard length 46-133 mm. Three specimens, 15 August 1964, tributary of Gran Rio
between Ligolio and Awaradam (Boeseman, RMNH no. 25713), standard length 49, 56,
143 mm. 34 specimens, 21 August 1964, creek below Moesoembaprati, Suriname Rivier
(Boeseman, RMNH no. 25714), standard length 47-166 mm. Four specimens, 20 March

1965, Suriname Rivier near Brokopondo (Mees, RMNH no. 25715), standard length 56,
132, 144, 160 mm. Two specimens, 10 May 1965, Mamadam, Suriname Rivier (Mees,
RMNH no. 25763), standard length 113, 137 mm. Three specimens, 12 May 1965, Suri¬
name Rivier near Botopassie (Mees, RMNH no. 25716), standard length 94, 100, 133 mm.
Eight specimens, 13 May 1965, Suriname Rivier near Botopassie (Mees, RMNH

10

146

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

no. 25717), standard length 46-125 mm. Two specimens, 14 May 1965, Mamadam, Suri¬
name Rivier (Mees, RMNH no. 25718), standard length 106, 125 mm. Six specimens,
16 July 1965, creek near Kajana, Gran Rio (Mees, RMNH no. 25719), standard length
35-44 mm. Two specimens, 18 July 1965, Grandam, Gran Rio (Mees, RMNH no. 25720),
standard length 62, 64 mm. One specimen, Gran Rio near Ligolio (Mees, RMNH
no. 25721), standard length 114 mm. Two specimens, 13 August 1965, Mamadam, Suri¬
name Rivier (Mees, RMNH no. 25722), standard length 805/2, 106 mm. Five specimens,
23 September 1965. Zandvallen, Kabalebo (Mees, RMNH no. 25723), standard length
102, 113, 118, 122, 130 mm. Eight specimens, 18 October 1965, Compagnie Kreek (Mees,
RMNH no. 25724), standard length 54-112 mm. One specimen, 17 November 1965, Tapa-
nahoni, ca. 2 km downstream from mouth of Paloemeu (Mees, RMNH no. 25725),
standard length 125 mm. Four specimens, 19 November 1965, Tapanahoni near mouth of
Paloemeu (Mees, RMNH no. 25726), standard length 57, 105, 107, 112 mm. Five speci¬
mens, 27 November 1965, Tapanahoni, ca. 2 km upstream from mouth of Paloemeu
(Mees, RMNH no. 25727), standard length 49, 53, 56, 57, 57 mm. 30 specimens, same
data (Mees, RMNH no. 25728), standard length 62-143 mm. One specimen, 30 November
1965, Tapanahoni, ca. 2 km downstream from mouth Paloemeu (Mees, RMNH no. 25729),
standard length 119 mm. Eight specimens, 12-14 December 1965, Langetabbetje, upper
course of Sara Kreek (Mees, RMNH no. 25730), standard length 55-82 mm. 73 speci¬
mens, 19 December 1965, Compagnie Kreek (Mees, RMNH no. 25731), standard length
44-121 mm. 49 specimens, 28 December 1965, Marshall Kreek (Mees, RMNH no. 25732),
standard length 50-117 mm. 20 specimens, 23 January 1966, Sipaliwini (Mees, RMNH
no. 25733), standard length 81-142 mm. Six specimens, same data (Mees, RMNH
no. 25734), standard length 525/2-67 mm. One specimen, 26 January 1966, Sipaliwini
(Mees, RMNH no. 25735), standard length 95 mm. Seven specimens, 6 February 1966,
Sipaliwini (Mees, RMNH no. 25736), standard length 73-137 mm. Eight specimens,
9 February 1966, Sipaliwini (Mees, RMNH no. 25737), standard length 66-138 mm.
One specimen, 5 March 1966, creek near Brokopondo (Mees, RMNH no. 25738), standard
length 75 mm. 65 specimens, 12 March 1966, tributary of Mama Kreek near Berg-en-Dal
(Mees, RMNH no. 25739), standard length 36-123 mm. 21 specimens, 15 March 1966,
Compagnie Kreek (Mees, RMNH no. 25740), standard length 63-122 mm.

Extra-limital material. One specimen ($), April 1864, Dtemerara (MV no. 45601),
standard length 61 mm, holotype of Pimelodus aliipinnis .

Characters. — D 1.6 (once I.7), A 11-14 (iii-iv.8-10), P I.9 or 1 .10, V 6
(1.5), C i.I5*i and rudiments (rarely only 13 or 14 divided rays), branchio-
stegals 6 or 7, gill-rakers 11-14 (2-3 + 1 + 7-10). Head 4.2-4.8, predorsal
length 3.35-3.4, depth of body 6.0-6.8, greatest width of body 5-7-6.8, base
of adipose fin 2.25-3.0 times in standard length; eye moderately large
(depending on sex as explained below), x.5-2.0 times in snout, 0.7-1.0 times
in bony interorbital, 3.5-4.5 times in head; maxillary barbels long, usually
reaching at least to the end of the adipose base, or to the caudal base, and
frequently to the tip of the caudal fin; dorsal spine not very strong, thin,
weakly serrated, its length 2.1-2.5 times in the predorsal length; pectoral
spines longer and stronger than the dorsal spine, along the anterior edge
weakly serrated, along the posterior edge with a series of strong recurved
hooks, placed dose together and varying in number (very small specimens
may have no more than six or seven; in large individuals this number may

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

147

increase to about 30), and not continued on the distal fourth of the spine
which is only weakly serrated or more or less smooth, their length 1.55-2.0
times in predorsal length.

There is a difference in relative size of the eyes between males and females
(confirmed by internal sexing); in some samples this is more distinct than
in others, but it may be said that generally speaking, in fishes of the size-
range here studied, specimens with eyes measuring less than four times in the
head are males, and specimens with eyes measuring over four times in the
head are females.

Colour remarkably pale, greyish-yellow, perhaps best described as sand-
colour, with on the flanks, following the lateral-line, a narrow black line
which varies geographically in development (see Discussion).

Size. Specimens of ca. 130 mm standard length and larger are sexually
mature. Eigenmann (1917: 231) gave: “a recorded length of 340 mm”
(probably total length). For a description of a large specimen from British
Guiana, see the Discussion.

Habitat and status. — This is easily the most abundant and most widely
distributed catfish of Suriname. Locally, in specialised habitats, it may
occasionally be outnumbered by other species, but on the whole no other
species approaches it in numbers. It is a bottom-fish, found in creeks as well
as in the rivers.

Distribution. — Found throughout Suriname. If my surmise that P.
cristata and P. gracilis are one species is correct, the range includes the
whole of tropical South America east of the Andes, from Venezuela to the
La Plata basin.

Discussion. — Specimens from the Marowijne basin (Tapanahoni) have
on an average the black lateral stripe wider than material from the Suriname
basin and more western localities. Eigenmann & Eigenmann (1890: 147,
key) mistakenly state that P . cristata has no dark lateral band: in fresh
material it is always present, but it may become indistinct in preservation.
Whether the Amazonian specimens on which the mentioned authors based
their description of P. cristata really belong to that species, remains to be
verified.

Mrs. McConnell has forwarded a single specimen from Karanambo, British
Guiana, June 1961, collected by Mr. McTurk, which is aberrant. Standard
length ca. 250 mm, D 1 . 6 , A iv.12, P 1 . 8 , V i.5, C 7 + 8 (divided rays),
branchiostegals 6, gill-rakers 3+1 + 7; adipose fin 89 mm, predorsal length
58 mm, eye 9 mm, 2.5 times in snout, 1.3 times in bony interorbital, and 5.2
times in head; maxillary barbels reaching to anal base; D-spine 1.5 times in
predorsal length; P-spines both with damaged tips, but probably they had been

148

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

a little shorter than D-spine; fins fleshy. The specimen is distorted, which
makes it impracticable to supply other proportions. The collector described
the colour of this fish as: “dark electric blue (almost black)". This fish is
considerably larger than any other specimen I have seen (the maximum size
found in almost a thousand specimens from Suriname is 176 mm standard
length), and its characters: the small eye, the fleshy fins, the slightly in¬
creased number of anal rays, and even the dark colour, may well be factors
of size. I do not believe that this specimen represents a different species,
although the possibility cannot be dismissed. Therefore it has not been
included in the list of material examined.

Van der Stigchel (1946: 64) pointed out that the fishes which Eigenmann
& Eigenmann (1890) described under the name Pimelodus altipinnis, do not
belong to that species as described by Steindachner (1864). He also sug¬
gested, following an observation made by Eigenmann (1912: 177), that the
type of Pimelodus altipinnis might be a Pimelodella and not a Pimelodus , but
made no further attempt to identify it. I have examined the holotype of
P. altipinnis , and found that van der Stigchel's surmise is perfectly correct:
the specimen is a typical Pimelodella, with a narrow occipital process meeting
the equally narrow dorsal plate, and I identify it as a female of P. cristata.
The specimen is from Demerara, April 1864, standard length 64 mm, total
length 78 mm, D 1 . 6 , A 12, P I.9, V 6, branchiostegals 7, adipose fin ca. 2.5
times in standard length, eye small, ca. 4.5 times in head, maxillary barbels
reaching to a little beyond base of caudal fin. On the accompanying label
the specimen was marked “Typus??”, but there is no need at all to doubt
that this is the type of P. altipinnis, as it agrees exactly with Steindachner’s
(1864) figure and description.

When Hoedeman (1961) described Pimelodella geryi , he compared it
with P. me galops only and made no mention of the common P . cristata.
Evidently he was misguided by Eigenmann’s (1912: 168) key, which gives
for P. cristata an adipose fin of 2 . 33 - 2.3 in the length, for P. megalops and
P. macturki an adipose fin of 3 or more in the length. In my large material
of P. cristata the length of the adipose fin varies from 2.25-3.0 times in
standard length (admittedly with the majority of specimens in the range
2.3~2.5). The type of P. geryi, a specimen that, incidentally, is in a very poor
condition, has an adipose fin which measures 3.0 times in standard length.
It falls therefore just in the established range of variation of P. cristata, and
as I have been unable to find other differences, I regard P. geryi as a syno¬
nym of P. cristata . The type of P. geryi has a conspicuous black lateral stripe,
as one would expect in P. cristata from the Marowijne basin.

Although I am not aware that this has ever been suggested, I consider

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

149

it very likely that P. cristata does not differ from P. gracilis (Valenciennes,
1840), which has nine years priority. This would also explain the great con¬
fusion between these names which one finds in the literature. The main differ¬
ence between the two nominal species appears to be in the shape of the caudal
fin, which is supposed to be much prolonged in P. gracilis, and not produced
in P. cristata.

From Eigenmann (1917: 231-232), the only author to have seriously
studied the genus in this century, I quote the following diagnoses as given
in his key:

“Lower caudal lobe much broader than the upper, upper lobe not produced,
the lower frequently the longer; lateral band if present narrow, linear; 13-14
teeth on the posterior face of the pectoral in specimens 83 mm long, about
20-30 much more feeble ones in large specimens, the more distal thorns
sometimes with a broad base and a long free outer edge; maxillary barbels

reaching origin, or beyond tip of adipose.

.2. cristata (Muller & Troschel)”.

“Upper caudal lobe normally greatly prolonged, but subequal in the young,
not much narrower than the lower lobe; maxillary barbel always extending
at least beyond base of anal; lateral band, if present, 'broad; pectoral spine a
little longer or a little shorter than the dorsal spine with a variable number

of strong teeth on over half its posterior margin.

.4. gracilis (Cuvier & Valenciennes)”.

The variation in length of the maxillary barbel in P. cristata is considerable,
many specimens having it reaching to or beyond the tip of the caudal fin.
Eigenmann’s (1917: pi- XXXV, fig. 37-41) own figures show that there is
a lot of individual variation in the pectoral spines, and that in general appear¬
ance those of P . gracilis and P. cristata agree. Variation in width of the
lateral band in P. cristata has already been discussed. As far as the shape of
the caudal fin is concerned, Eigenmann himself made qualifications (“sub¬
equal in the young”), and my large material of P. cristata shows that there
is a lot of variation in this character. The normal condition in P. cristata
is that the upper lobe of the caudal fin is a little longer, but more slender than
the lower lobe (the usual number of developed caudal rays is 8 + Q, or
branched rays 7 + 8), and in a few specimens the upper lobe is considerably
elongated, although in others the lower lobe is actually the longer. From Uru¬
guay, where P. cristata is not known to occur, Devincenzi & Teague (1942)
and Devincenzi & Barrattini (1926-1940: pi. XLV) describe and figure under
the name of P. gracilis a fish with symmetrically forked tail. Ringuelet, Aram-
buru & de Aramburu (1967: 306) believe to have found another character
to differentiate the two species: adipose fin 2.2 in standard length (P.

150 ZOOLOGISCHE VERHANDELINGEN 132 (l 974 )

cristata) and adipose fin 2.3-3.0 in standard length ( P . gracilis ) *). In speci¬
mens of P. cristata from Suriname the adipose fin is 2.25-3.0 in standard
length.

Whatever the identity of P. gracilis, there is little doubt that specimens
recorded under this name from British Guiana (Fowler, 1914) and the
Orinoco (Steindachner, 1879; Schultz, 1944a), are the same as P. cristata
(note that the fish figured by Fowler, 1914, fig. 13, under the name P.
gracile shows an ordinary forked tail, without produced upper lobe). If
P . gracilis does differ from P. cristata , the former may be confined to the
Uruguay-Parana-La Plata basins, as suggested by Eigenmann (1917), who
provided all localities outside that region with a query. For Amazonian speci¬
mens a third name is available, P. steindachneri, launched with much hesitation
by Eigenmann (1917).

In view of existing uncertainties, I have limited the synonymy and biblio¬
graphy to records from the Guianas, which are the only ones certainly
referable to P. cristata.

I take this opportunity to point out that as type-locality of P. gracilis ,
Buenos Aires is usually quoted (cf. Fowler, 1951: 540), but that this is an
error, as reference to the original description (Valenciennes in Cuvier &
Valenciennes, 1840: 134-135) shows; I quote:

“M. d’Orbigny a encore envoye de Buenos-Aires une espece assez voisine
des trois precedentes, et qui a la meme longue adipeuse, ... M. d’Orbigny
les a pris dans la province de Corrientes dans le Parana et les autres rivieres
au-dessus de 28° de latitude sud, toujours au milieu des courans dans les
lieux pierreux”.

From this it is evident that Buenos Aires was only the place of dispatch of
the specimens, and that the actual type-locality is the Rio Parana in Corrien¬
tes, the only locality definitely mentioned by Valenciennes.

Pimelodella macturki Eigenmann

Pimelodella macturki Eigenmann, 1912, Mem, Carnegie Mus., 5: 170, pi. XVI fig. 1 —
Creek in Mora Passage, British Guiana.

Pimelodella macturkii Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3:
389 — nomen nudum (Coastal streams of British Guiana).

Pimelodella macturki; Eigenmann, 1917, Mem. Carnegie Mus., 7: 248 (British Guiana
near the coast); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 45 (Guiana
Inglesa); Bohlke, 1953, Stanford Ichth. Bull., 5: 44 (Choca trenches, British Guiana) ;
Lundberg & Baskin, 1969, Amer. Mus. Novit., 2398: 6, 29 (no locality).

Pimelodella macturkii ; Henn, 1928, Ann. Carnegie Mus. 19: 77 (Creek in Mora Pas¬
sage) ; van der Stigchel, 1946, South American Nematognathi: 57 (Surinam; Morawhana,

1) In the key in this book, under no. 2, clearly “cuerpo” should be read for “cabeza”.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE 151

British Guiana); van der Stigchel, 1947, Z00L Meded, 27: 57 (Surinam; Morawhana,
British Guiana); Grey, 1947, Fieldiana, Zool., 32: 174 (British Guiana: Mora Passage).

Material. — One specimen, 1902/1910, Suriname (D. G. J. Bolten, RMNH no. 17254),
standard length 70 mm 1 ).

Characters. — D 1 . 6 , A 14 (v.9), P 1 . 8 , V 6 (i.5), C i.is.i and rudiments,
'branchiostegals 8, gill rakers 3 + 1+8. Head 4.1, predorsal length 2.95,
depth of body 4.1, greatest width of body (between the cleithra) 5.3, base
of adipose fin 3.8 times in standard length; eye large, 1.1 times in snout, 0.8
times in bony interorbital, 3.5 times in head; maxillary barbels relatively
short, reaching to a vertical through the anus; dorsal spine not very strong,
weakly serrated on both edges, its length 2.5 times in the predorsal length;
pectoral spines along the anterior edge with weak serrations, along the
posterior edge with 12 strong recurved hooks, not continued on the distal
fourth of the spine, which is smooth, its length 2.1 times in predorsal length.
P. macturki differs from P. cristata mainly by the much shorter adipose fin,
and the shorter maxillary barbels.

Distribution. — On present evidence the range of this species is very
restricted; it occurs in the coastal region from Demerara east to, probably,
the vicinity of Paramaribo.

Discussion. — The inclusion of C . macturki in the fauna of Suriname
rests on the single specimen listed above, which had previously been studied
and recorded by van der Stigchel (1946). In view of the faunistic impor¬
tance of this specimen and its somewhat unsatisfactory label, I have checked
if its provenance is above suspicion. This appears to be the case: the collec¬
tor, Dr. D. G. J. Bolten (born’s Gravenhage, 5 April 1871 — deceased Zeist,
29 Aug. i960) lived in Paramaribo for a considerable time (Oct. 1902-May
1910) and he presented to our museum a collection of 186 fishes, besides
other zoological specimens, in 1910 (cf. Jentink, 1910: 15). The date given
on the label (not an original one) and published by van der Stigchel, June
1910, is, however, undoubtedly not the date of collecting, but the month in
which the specimen was received in Leiden. Bolten has personally collected
his material, and this was done in eastern coastal Suriname (cf. Holthuis,
1959: 27), throughout the period of his stay. In the light of this evidence,
there is in my opinion no reason to query the authenticity of the record.

Eigenmann’s (1912: 91, 170) experience in British Guiana was that this
species is confined to coastal streams, where it takes the place of P. cristata,
which occurs more inland. Undoubtedly the situation in Suriname is similar,

1) Tn March and April 1971, Boeseman and Geijskes collected two samples of Pitnelo-
della macturki in tributaries of the Maratakka, Nickerie District: a welcome confirmation
of the occurrence of the species in Suriname.

152

ZOOLOGISCHE VERHANDELINGEN 1 32 (l974)

and this explains why P. macturki was not obtained by us, as our collecting
activities were almost confined to inland localities.

Pimelodella boschmai van der Stigchel

Pimelodella boschmai van der Stigchel, 1964 (17 Jan.), Zool. Meded., 39: 327 — Mogi-
Guassu (river) below the Cachoeira de Emas (Emas falls).

Pimelodellla insignis Schubart, 1964 (21 May), Bol. Mus. Nac. Rio de Janeiro, Zool.,
244: 6 — Rio Mogi Guagu.

Discussion. — Notwithstanding a single discrepancy — van der Stigchel
(1964) stressed the fact that it is not the filament of the dorsal spine, but
the first dorsal ray which is elongated, whereas Schubart (1964) described
and figured an elongated filament to the spine, and a first ray which is not
produced — P. boschmai and P. insignis are clearly one species. Van der
Stigchel mentioned that he received his specimens from Schubart, and it is
apparent that the two nominal species were based on material originally
belonging to the same samples.

The type-material of P. boschmai (RMNH nos. 23248, 23250, 23251)
shows a dorsal fin as described by its author, with he first ray, and not the
spine, produced. A plausible solution of this discrepancy might be that both
the filament of the spine and the first ray can sometimes be elongated; note
that Schubart shows in his table that several of his specimens had a first
ray nearly as long, in one case out of ten even longer, than the “bandeira”.

As P. boschmai was published a few months before P. insignis, the former
name has priority. The species appears to be very close to P. griff ini Eigen-
mann.

Genus Rhamdia Bleeker

Rhamdia Bleeker, 1858, Visschen Ind. Arch., 1: 197 — type by original designation,
Pimelodus sebae Valenciennes.

Pimelenotus Gill, 1858, Ann. Lyc. Nat. Hist. New York, 6: 387 — based on a series
of species now placed in the genera Rhamdia and Pimelodella ; type by subsequent
selection (Eigenmann & Eigenmann, 1890: 116), Pimelenotus Vilsoni Gill (— Rhamdia
qaelen ) *).

Caecorhamdia Norman, 1926, Ann. Mag. Nat. Hist., (9) 18: 325 — type by original
designation and monotypy, Caecorhamdia uric hi Norman.

Generic diagnosis. — In general appearance very close to Pimelodella, but
the occipital process, although in most species well developed, does not reach
the dorsal plate.

1) Rhamdia Bleeker, 1858, is generally accepted as having priority over Pimelenotus
Gill, 1858, but I do not know on what this is based. Eigenmann & Eigenmann (1890: 116)
give 1859 as year of publication of the genus Pimelenotus but on page 126 quote
P. Vilsoni as having been published in 1858.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

153

Distribution. — A very widely-distributed genus, occurring throughout
Central America to as far north as Mexico, and in northern and eastern
South America to the Rio de la Plata; in western South America even high
in the Andes (Lake Titicaca), but apparently not found in rivers draining
into the Pacific. From Suriname only one, widely distributed, species is known,
which is common.

Discussion. — The genus is assumed to be a large one: Gosline (1945)
listed 66 forms (57 species and 9 subspecies), not all of these valid, and
several additional species have been described in more recent years.

The grounds for not recognizing Caecorhamdiu will be given in the dis¬
cussion of Rhamdia quelen urichi.

Rhamdia quelen (Quoy & Gaimard) (fig. 33)

Pimelodus quelen Quoy & Gaimard, 1824, Voy. Uranie, Zool.: 228, pi. 49, fig. 3, 4 —
Montevideo (reference not verified).

Heterobranchus sextentaculatus Spix, 1829, Gen. Spec. Piscium: 28, pi. XI — in
Brasiliae aequatorialis fluviis.

Pimelodus Sebae Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.), IS: 125 — Surinam, Cayenne, Rio-Janeiro, Buenos-Ayres.

Pimelodus Sellonis Muller & Troschel, 1849, Horae Ichth., 3:2 — Brasilien.

Pimelodus Stegelichii Muller & Troschel, 1849, Horae Ichth., 3:3 — Surinam.

Pimelodus muscuius Muller & Troschel, 1849, Horae Ichth., 3:4 — America.

Pimelotus (misprint for Pimelenotus, see errata at end of volume) Vilsoni Gill, 1858,
Ann. Lyc. Nat. Hist. New York, 6: 391 — Trinidad.

Pimelodus holomelas Gunther, 1863, Ann. Mag. Nat. Hist., (3) 12: 442 — the
Essequibo.

Pimelodus miilleri Gunther, 1864, Cat. Fish. Brit. Mus., 5: 119 — River Capin, Para;
Para; Surinam.

Pimelodus zvuchereri Gunther, 1864, Cat. Fish. Brit. Mus., 5: 123 — Bahia.

[Pimelodus (Rhamdia) Queleni) var. cuprea Steindachner, 1876, Sitzb. Akad. Wiss.
Wien, Mathem.-Naturw. Cl., 74 (1) : 623 — Parahyba.

Pimelodus (Rhamdia) Knerii Steindachner, 1876, Sitzb. Akad. Wiss. Wien, Mathem.-
Naturw. Cl., 74 (1) : 631, Note —- Marabitanos.

Pim\elodus] Cuyabae Steindachner, 1876, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw.
Cl., 74 (1): 633, Note —- Cuyaba.

Pimelodus Sebae ; Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw. Cl., 26
(1857): 417 (Marabitanos); Hyrtl, 1859, Denkschr. Akad. Wiss. Wien, Mathem.-Naturw.
Cl., 16 (1) : 5, 15, 16 (no locality) ; Goeldi, 1898, Bol. Mu&. Paraense, 2: 461 (Rio Branco).

Rhamdia Sebae ; Bleeker, 1838, Visschen Ind. Arch., 1: 207 (Amer. merid.).

Rhamdia Queleni ; Bleeker, 1862. Atlas Ichth., 2: 11 (name only); Bleeker, 1863, Ned.
Tijdschr. Dierk., 1: 101 (name only); Bleeker, 1864, Nat. Verh. Holl. Maatsch. Weten-
sch., (2) 20: 73 (Surinama).

Pimelodus sebae ; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 119 (Demerara; British
Guiana; Brazils; South America); Boulenger, 1900, Bol. Mus. Torino, 15 (370): 1
(Urucum et Carandasinho, pres de Corumba); Regan, 1905, Proc. Zool. Soc. Lond.: 190
(Rio Negro) ; Puyo, 1949, Poiss. Guyane Fr.: 90 (Guyane Franqaise: les criques vaseuses
tributaires de la Mana et du fleuve Kourou).

Pimelodus holomelas ; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 120 (Essequibo);
Vaillant, 1898, Notes Leyden Mus., 20: 3, 5 (New Amsterdam, Berbice).

154

ZOOLOGISCHE VERHANDELINGEN I 32 (1974)

Pimelodns stegelichii ; Gunther, 1864, Cat. Fish. Brit Mus., 5: 121 (Demerara;
Surinam).

Pimelodus queleni ; Gunther, 1864, Cat Fish. Brit. Mus., 5: 123 (Brazil); Boulenger,
1900, Bol. Mus. Torino, 15 (370) : 1 (Urucum et Carandasinho, pres de Corumba).

Pimelodus (Rhamdia) Queleni ; Steindachner, 1876, Sitzb. Akad. Wiss. Wien, Mathem.-
Naturw. CL, 74 (1): 622 (Rio Parahyba bei Iuiz de Fora, Campos; Rio doce, Porto
alegre, Cannavierias; Amazonenstrom bei Para; Bahia).

Pimelodus (Rhamdia) Sebae ; Steindachner, 1876, Sitzb. Akad. Wiss. Wien, Mathem.-
Naturw. CL, 74 (1) : 626 (vom Demerara, aus dem Essequibo und von St. Martha zu-
nachst der Miindung des Magdalenenstromes); Steindachner, 1879, Denkschr. Akad.
Wiss. Wien, Mathem.-Naturw. CL, 39 (1) : 32 (Magdalenen-Strom); Steindachner,
1880, Denkschr. Akad. Wiss. Wien, Mathem.-Naturw. CL, 42 (1) : 59 (Cauca).
Pimelodus Mulleri ; Kappler, 1887, Surinam: 152 (Surinam).

Rhamdia sebae ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2) 1: 126
(Tonatins; Gurupa; Rio de Janeiro; Bahia; Xingu; Santa Cruz; Cudajas; Sao Matheos;
Rio Doce; Serpa; Tabatinga; Goyaz; Para; Teffe; Surinam; Villa Bella) ; Eigenmann &
Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1: 123 (localities as before); Eigenmann,
1891, Proc. U.S. Nat Mus., 14: 28 (Rio de Janeiro to Rio Magdalena; Amazon; Soli-
moens) ; A. de Miranda Ribeiro, 1907, A Lavoura, 11: 186 (Barra de Pariqueragu, Rio
da Ribeira); Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3 : 385 (Rio de
Janeiro and Paraguay to Rio Magdalena and Tabatinga); A. de Miranda Ribeiro, 1911,
Arch. Mus. Nac. Rio de Janeiro, 16: 279 (many localities); Eigenmann, 1912, Mem.
Carnegie Mus., 5: 164 (several localities in British Guiana); Starks, 1913, Stanford Publ.
Univ. Ser.: 27 (Para); A. de Miranda Ribeiro, 1914, Comm. Linh. Telegr. Estr. Matto-
Grosso ao Amazonas, 15, Annexo 5: 5 (Rio Manso; S. Manoel; Manaos); Fowler, 1914,
Proc. Acad. Nat. Sci. Philad., 66: 258 (Rupununi River, British Guiana); Bertoni, 1914,
Fauna Paraguaya: 8 (Paraguay); Fowler, 1915, Proc. Acad. Nat Sci. Philad., 67: 209
(Surinam and Peruvian Amazon) ; Popta in Benjamins & Snelleman, 1914-1917, Encycl.
Ned. West-Indie: 607 (Suriname); Eigenmann & Fisher, 1917, Ann. Carnegie Mus., 11:
396 (Entre Rios, Rio Parahyba; Campos, Rio Parahyba; Barra da Pirahy, Rio Parahyba;
Jacarehy, Rio Parahyba; Rio das Velhas; San Joaquin; Para); A. de Miranda Ribeiro,
1918, Rev. Mus. Paul., ro: 729 (Taubate; Rio Grande do Sul; Rio Doce, E. Santo;
Iguape; Itaborahy); Beebe, 1923, Zoologica, 6: 118 (Kartabo); Pearson, 1937, Proc.
Calif. Acad. Sci., (4) 23: in (Mamore and Paraguay basins); Fowler, 1940, Proc. Acad.
Nat. Sci. Philad., 91: 222 (Ucayali River basin, Peru); Fowler, 1940, Proc. Acad. Nat.
Sci. Philad., 92: 95 (San Joaquin, Bolivia); Fowler, 1941, Bol. Mus. “Javier Prado”, 5:
370 (Contamana, Peru) ; Giintert, 1942, Zool. Anz., 138: 31 (Nariva River auf Trinidad;
Rio Tocuj^o bei Boca Tocuya in Venezuela); Eigenmann & Allen, 1942, Fish. W. South
America: 94 (Magdalena basin to that of Rio Janeiro; throughout the Amazon) ; Fowler,
1943, Proc. Acad. Nat. Sci. Philad., 95: 244 (Villavicencio and Rio Ocoa, Colombia);
Schultz, 1944, Proc. U.S. Nat. Mus., 94: 194 (many localities in the Maracaibo Basin
of Venezuela); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Z00L, 33: 36 (do Rio de
Janeiro ao rio Madalena e Tabatinga); Fowler, 1945, Les Peces del Peru: 38 (Conta¬
mana, Peru) ; Sterba, 1959, Siisswasserfische aus aller Welt: 306, fig. 647 (Im ganzen
nordlichen und mittleren Siidamerika ostlich der Anden) ; Boeseman, i960, Studies Fauna
Curagao, 10: 100 (along Talparo Road, Trinidad); Vogt, 1970, Grzimeks Tierleben, 4:
403 (Siidamerika).

Rhamdia sebae kneri ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2) 1:
126 (Tabatinga, Jutahy); Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci.,
1: 126 (Tabatinga and Jutahy); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus.,
14: 28 (Amazon, Solimoens, and northward); Eigenmann, 1910, Rep. Princeton Univ.
Exp. Patagonia, 3: 385 (Upper Paraguay and northward) ; Pearson, 1937, Proc. Calif.
Acad. Sci., (4) 23: in (Paraguay basin); Giintert, 1942, Zool. Anz., 138: 32 (Nariva

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

I SS

River auf Trinidad); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Z00L, 33: 36 (alto
Paraguai e para o norte).

Rhamdia quelen ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2) 1: 126
(Santa Clara; Rio Mucuri; Iuiz de Fora; Campos; Rio Jequitinhonha; Mendez; Rio de
Janeiro; Macacos; Sao Matheos; Rio Parahyba; Cannavierias; Rio Grande do Sul);
Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1: 127 (localities as
preceding reference); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 28
(La Plata to Amazon); Eigenmann & Norris, 1900, Rev. Mus. Paul., 4: 350 (Taubate);
Eigenmann & Kennedy, 1903, Proc. Acad. Nat. Sci. Philad., 55: 499 (Estancia La
Armonia; Asuncion and Campo Grande, Paraguay); Eigenmann & Bean, 1907, Proc.
U.S. Nat. Mus., 31: 660 (Amazon River between Para and Manaos) ; Eigenmann, 1907,
Proc. Wash. Acad. Sci., 8: 450 (near Buenos Aires); Eigenmann, McAtee & Ward, 1907,
Ann. Carnegie Mus., 4: 113 (Corumba, Paraguay); Eigenmann, 1910, Rep. Princeton
Univ. Exp. Patagonia, 3: 386 (Rio de la Plata to the Amazon and Guiana); A. de
Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 278 (lagos da Quinta Boa
Vista em S. Christovao); Eigenmann, 1912, Mem. Carnegie Mus., 5: 163 (several
localities in British Guiana); Starks, 1913, Stanford Publ. Univ. Ser.: 27 (Ceara Mirim);
A. de Miranda Ribeiro, 1914, Comm. Linh. Telegr. Estr. Matto-Grosso ao Amazonas, 15,
Annexo 5: 4 (Coxipo da Ponte); Bertoni, 1914, Fauna Paraguaya: 8 (Alto Parana);
Fowler, 1915, Proc. Acad. Nat. Sci. Philad., 67: 21 x (eastern Brazil and the La Plata);
Eigenmann & Fisher, 1917, Ann. Carnegie Mus., 11: 396 (many localities in south¬
eastern Brazil and the Paraguay basin); A. de Miranda Ribeiro, 1918, Rev. Mus. Paul.,
10: 729 (Rio Parahytinga; Hansa; Sorocaba; Mercado de S. Paulo; S. Paulo; Lagoa
Feia, Rio de Janairo; Itanhaem do Bananal, Parana Mirim) ; Devincenzi, 1924, An. Mus.
Nac. Montevideo, (2) 1: 149 (Uruguay: several localities) ; Pearson, 1924, Indiana Univ.
Stud., 11 (64) : 11 (Rio Beni basin, Bolivia: Tumupasa, Lake Rogoagua, Espia) ; Myers,
1930, Proc. Biol. Soc. Wash., 43: 69 (Rio Meta, Colombia) ; Magalhaes, 1931, Mon.
Brazil. Peix. Fluv.: 130, fig. 58 (rios do Estado de S. Paulo, Rio de Janeiro e Parana);
Pearson, 1937, Proc. Calif. Acad. Sci., (4) 23: in (Beni-Mamore and Paraguay basins) ;
Fowler, 1940, Proc. Acad. Nat. Sci. Philad., 92: 95 (Bolivia: Puerto Suarez; Tumpasa;
L. Rogoagua; Espia); Eigenmann & Allen, 1942, Fish. W. South America: 93 (Yuri-
maguas; Iquitos; La Merced; Puerto Melendez, Pongo de Manseriche); Fowler, 1943,
Proc. Acad. Nat. Sci. Philad., 95: 318 (Cerro Largo, Montevideo); Schultz, 1944, Proc.
U.S. Nat. Mus., 94: 194, 196 (British Guiana; Valencia Basin of Venezuela) ; Gosline,
1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 36 (do Rio da Prata a Guiana e Vene¬
zuela) ; Fowler, 1945, Los Peces del Peru: 39 (Peru: Yurimangas, Iquitos, La Merced,
Puerto Melendes, Pongo de Manseriche); van der Stigchel, 1946, South American Ne-
matognathi: 48 (Surinam; Berbice River, New Amsterdam, British Guiana; Rio de
Janeiro; Brazil; swamps behind the Botanic Garden, Lower Surinam River; Upper
Saramacca; Port Real, Rio de Janeiro; Aruka River, British Guiana; Aruataima Falls,
Upper Potaro River, British Guiana); van der Stigchel, 1947, Zool. Meded., 27: 50 (as
preceding reference); de Buen, 1950, Publ. Cient. S.O.Y.P., 2: 71 (Uruguay); Bertin &
Esteve, 1950, Cat. Types Poiss. Mus. Paris, 5: 56 (Cayenne (Guyane); Buenos-Ay res
(Argentine); FI. Guyayaquil (Equateur)); Boeseman, 1952, Zool. Meded., 31: 181
(Surinam: Coropina creek, Republiek; Nanni creek, Doksen savanna; Surinam River,
Paramaribo) ; Boeseman, 1953. Zool. Meded., 32: 6 (Surinam: Sommelsdijkse Kreek, near
Wolfenbuttel (asylum), Paramaribo; Langaman kondre (near Albina); Wanekreek;
Djaikreek; Taproepa Creek; Marowini Basin, Nassau Mountains); Inger, 1956, Fiel-
diana, Zool., 34: 425, 428 (Rio Abacapa, Venezuela); Boeseman, i960, Studies Fauna
Curasao, 10: 100 (along Talparo Road, Trinidad); Luengo, 1963, Acta Biol. Venez., 3:
330 (references only); Fernandez-Yepez, 1967, Acta Biol. Venez., 5: 161 (Venezuela: rio
Uruyen medio); Ringuelet, Aramburu & de Aramburu, 1967, Los Peces Argentinos de
Agua Dulce: 322 (distibution).

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

156

Rhamdia Quelen; Berg, 1895, An. Mus. Nac. Buenos Aires, 4: 133 (Argentina: Arroyo
del Tala de Catamarca; Rio Parana, cerca de San Nicolas).

Pimelodus Miilleri; Goeldi, 1898, Bol. Mus. Paraense, 2: 459 (Magoary-Marajo;
Counany).

Pimelodus Muelleri ; Goeldi, 1898, Bol. Mus. Paraense, 2: 476 (Counany; Marajo).

Pimelodus muelleri; Regan, 1905, Proc. Zool. Soc. Lond.: 190 (Rio Negro).

Pimelodus (Rhamdia) wilsoni; Regan, 1906, Proc. Zool. Soc. Lond.: 386 (Trinidad).

Rhamdia vilsoni ; Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 385
(Trinidad); Fowler, 1915, Proc. Acad. Nat. Sci. Philad,, 67: 209 (Trinidad); Fowler,
1915, Proc. Acad. Nat. Sci. Philad., 67: 529 (Trinidad); Gosline, 1945, Bol. Mus. Nac.
Rio de Janeiro, Zool., 33: 36 (Trinidad).

Rhamdia holomelas ; (?) Eigenmann, 1912, Mem. Carnegie Mus., 5: 166 (Lama Stop-
Off, British Guiana) ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 36 (Esse-
quibo).

Rhamdia queleni; Norman, 1926, Ann. Mag. Nat. Hist., (9) 18: 325, 326, 327 fig. 2
(eastern South America north of the Rio Plata) ; Arnold & Ahl, 1936, Fremdl. Siiss-
wasserfische: 252, fig. (im siidlichen Brasilien); Devincenzi & Teague, 1942, An. Mus.
Montevideo, (2) 5 (4) : 43 (Uruguay) ; Sterba, 1959, Siisswasserfische aus aller Welt:
306, fig. 644 (Sudliches Brasilien, La-Plata-Gebiet).

Rhandia quelen ; Campos, 1944, Pap. Avuls. Zool., 4: 172 (Est. de Sao Paulo); Schu-
bart, 1964, Bol. Mus. Nac Rio de Janeiro Zool., 244: 18 (Rio Mogi Guagu).

Rhamdia sapo ; (pt.) van der Stigchel, 1946, South American Nematognathi: 52 (Para¬
maribo, Surinam); (pt.) van der Stigchel, 1947, Zool. Meded., 27: 52 (Paramaribo,
Surinam).

Pimelodus sehoe; Puyo, 1949, Poiss. Guyane Fr.: 87, 90 (Guyane Frangaise: dans les
criques vaseuses tributaires de la Mana et du fleuve Kourou).

Rhamdia sebae knerii; Fowler, 1951, Arq. Zool. S. Paulo, 6: 576 (Amazonas, Alto
Paraguay, Paraguay, Guiana).

Rhamdia sebae sebae ; Fowler, 1951, Arq. Zool. S. Paulo, 6: 577 (distribution).

Pimelodclla gracilis ; Price, 1955, J. Agric. Soc. Trinidad & Tobago, Soc. Pap., 863: 14
(Trinidad).

Pimelodella chagresi ; Price, 1955, J. Agric. Soc. Trinidad & Tobago, Soc. Pap., 863:
14 (Trinidad).

Pinirampus pinirampu; Hoedeman, 1961, Bull. Aquatic Biol., 2: 135 (French Guiana:
Litany river, village Aloike).

Material. — 88 specimens as follows. One specimen, 27 November 1957, Litani R. near
Alowike (J. Gery, ZMA no. 102234), standard length ca. 127 mm. One specimen, 29
January 1964, Cederkreek near Brokopondo (Boeseman, RMNH no. 25741), standard
length 162 mm. One specimen, 24 February 1964, Amanipani Kreek, a tributary of the
Sara Kreek between Locus and Dam (Boeseman, RMNH no. 25742), standard length
161 mm. One specimen, 6 March 1964, Gran Kreek, ca. 4 km from its mouth (Boeseman,
RMNH no. 25743), standard length 147 mm. Four specimens, 10 July 1964, shore of
Brokopondo Meer, between Kabel and Abontjima (Boeseman, RMNH no. 25762),
standard length 47, 99, 137, 156 mm. 20 specimens, 21 July 1964, tributary of Gran Kreek,
ca. 12 km from its mouth (Boeseman, RMNH no. 25744), standard length 28-62^ mm.
One specimen, 27 July 1964, shore of Brokopondo Meer between Kabel and Brownsweg
(Boeseman, RMNH no. 25745), standard length 76 mm. Four specimens, 30 July 1964,
tributary of middle course Gran Kreek (Boeseman, RMNH no. 25746), standard length
53 , 67, 69, 74 mm. Three specimens, 15 August 1964, tributary of Gran Rio between
Ligolio and Awaradam (Boeseman, RMNH no. 25766), standard length 53, 74, 88 mm.
Two specimens, 30 March 1965, Brokopondo Meer between Kabel and Brownsweg (Mees,
RMNH no. 25747), standard length 76, 93 mm. One specimen, 2 April 1965, Gansee,

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

157

Brokopondo Meer (Mees, RMNH no. 25748), standard length 196 mm. One specimen,
same data (Mees, RMNH no. 25749), standard length 120 mm. One specimen, 8 April
1965, Gansee, Brokopondo Meer (Mees, RMNH no. 25749), standard length 212 mm.
One specimen, 5 September 1965, Compagnie Kreek (Mees, RMNH no. 25750), standard
length 78 mm. One specimen, 27 November 1965, Tapanahoni, ca. 2 km upstream from
Paloemeu mouth (Mees, RMNH no. 25761), standard length xoi mm. Three specimens,
12-14 December 1965, Langetabbetje, upper course Sara Kreek (Mees, RMNH no. 25751),
standard length 118, 121, 185 mm. 13 specimens, 19 December 1965, Compagnie Kreek
(Mees, RMNH no. 25752), standard length 102-211 mm. Six specimens, 28 December

1965, Marshall Kreek (Mees, RMNH no. 25753), standard length 87, 92, 106, 122, 128,
137 mm. Three specimens, 26 January 1966, Sipaliwini (Mees, RMNH no. 25754),
standard length 54, 55, 100 mm. One specimen, 28 February 1966, Makambie Kreek near
Brownsweg (Mees, RMNH no. 25755), standard length 143 mm. Two specimens, 5 March

1966, creek behind Hermansdorp, Brokopondo (Mees, RMNH no. 25756), standard length
1 50, 157 mm. One specimen, 8 March 1966, Tapoeripa Kreek near Brokopondo (Mees,
RMNH no. 25757), standard length 157 mm. 16 specimens, tributary of the Mama Kreek
near Berg-en-Dal (Mees, RMNH no. 25758), standard length 104-150 mm.

Characters. — D 1 . 6 , A 10-13 (usually iii.8), P 1.8 or I.9, V 6 (i.5), C 15
or 16 (branched rays only), branchiostegals 7, gill-rakers 8-12 (2 or 3 + 1 +
5-8), adipose fin long, 24-2.7 in standard length. Superficially this species
is very similar to Pimelodella cristata] besides the difference in development
of the processus occipitalis it can be distinguished by its smaller eyes (5.6-74,
but usually 6 or more times in head), more depressed snout, and colour which
is dark brown in life and in freshly preserved material, against pale yellowish
in P. cristata.

Distribution. — In Suriname found throughout the country, in the same
habitat as Pimelodella cristata, but less numerous, although still a very com¬
mon species. Widely distributed in tropical South America, including Trini¬
dad; closely related forms in Central America.

Discussion. —• There has been much uncertainty about the validity of
R. sebae. Valenciennes (1840: 128) himself suggested the synonymy of R.
sebae with R. quelen, and he was followed by Bleeker (1862). Steindachner
(1876b: 626), however, disagreed and came with the character of the barbel-
length, that has since been widely used to distinguish R . sebae from R. quelen,
amongst others by the Eigenmanns (1890). Starks (1913: 27), although he
listed the two as different species, did not believe in the mentioned character
and considered it probable that they represented only one species. No progress
was made until van der Stigchel (1946) measured a number of specimens
on the basis of which he concluded definitely that R . sebae is a synonym of
R. quelen . He was followed “provisionally” by Boeseman (1952, 1953), but
later the same author (Boeseman, i960) reverted to the use of both names.
Boeseman’s argument when making this change of stand was that van der
Stigchel had expressed “an opinion not yet shared by any other authors”,

1 58 ZOOLOGISCHE VERHANDELINGEN 132 (1974)

but I am not aware that anybody since van der Stigchel has seriously studied
the species.

The position appears to be very much the same as that of Pimelodella
gracilis and P. cristata. Both the names R. quelen and R. sebae have been
applied widely over the whole of eastern and northern South America. In
Suriname, however, certainly no more than one species of Rhamdia occurs,
which in its individual variation includes all the characters ever used to
distinguish two species.

Therefore I agree with van der Stigchel that R. sebae is not specifically
different from R. quelen . I want to leave open the possibility that careful
measurement of large series from the La Plata and from Suriname would
reveal slight differences that in future could be used for the recognition of
subspecies within the huge range of the species.

In this connection I want to point out that subspecific names have been
applied in the species (viz., Rhamdia sebae kneri), with a complete lack of
understanding of the nature of subspecies, a matter I have discussed at some
length under Pseudoplatystoma fasciatum.

The reason that in this case I go further than in that of Pimelodella gracilis
and P. cristata , is that van der Stigchel's conclusion was based on an actual
comparison of material, whereas I have been unable to examine adequate
topotypical material of P. gracilis.

It is likely that several other nominal species belong in the synonymy of
R. quelen. As an example I can mention Rhamdia lehmanni Dahl (1961),
which was described without any reference to other members of its genus,
whereas in its description no characters are given that would differentiate it
from R. quelen.

In the original description of P. holomelas no mention is made of R. quelen
(or its synonym R. sebae), and no distinguishing characters from R. quelen
are given. The name was placed in the synonymy of Rhamdia sebae by Eigen-
mann & Eigenmann (1890) but resurrected without reference to his earlier
opinion by Eigenmann (1912), who distinguished R. holomelas from R.
quelen and R. sebae by having “Premaxillary band of teeth with a backward
projecting angle laterally. Maxillary barbel not extending to middle of
adipose; space between the eyes 3 in the head; caudal cleft, both lobes rounded,
the third or fourth ray from the cleft in the upper lobe longest; adipose dorsal
2-2.4 i n th e length”. Eigenmann’s material consisted of seven specimens from
Lama Stop-Off, British Guiana, whereas he stated also to have examined the
types in the British Museum. In the description he further states: “maxillary
barbel extending beyond the base of the pectoral, but not to its tip” — which
is surprising as in the original description Gunther wrote: “Maxillary barbels

MEES, AUCHENIPTERIDAE AND PIMELODIDAE 159

extending beyond the commencement or to the middle of the adipous fin.”
Subsequently van der Stigchel (1946: 52) commented on Eigenmann’s (1912)
work, and once more synonymized P. holomelas with R. quelen. I follow
van der Stigchel, but want to leave open the possibility that the specimens
described by Eigenmann (1912) are referable to a different species, as their
maxillarly badbels appear to be remarkably short. On the other hand, the speci¬
mens which Eigenmann identified as Rhamdia holomelas were rather large
(275-345 mm) and there are indications that in R. quelen larger individuals
have relatively shorter barbels than smaller individuals (cf. the large speci¬
mens discussed in the following paragraph).

Van der Stigchel (1946: 52-53) has listed another member of the genus
from Suriname, Rhamdia sapo (Valenciennes). As all previous records of
that species are confined to the Uruguay-Parana-Paraguay river systems, this
means an enormous extension of range. Although the examples of several
other species show that such a wide range is not impossible, it appeared
nevertheless desirable to verify the record, especially as van der Stigchel did
not comment on it, and does not appear to have realised that it was
unusual *).

The specimens (ZMA no. 100065) are distinguishable from the available
material of R. quelen mainly by their large size (standard lengt 313, 348 mm,
against a maximum of 212 mm in R. quelen). The maxillary barbels are
rather short, reaching a little beyond a vertical through the dorsal origin.

Information on the labels says that these specimens came from Paramaribo,
Suriname, and were presented to “Artis” by Bolten (obviously Dr. D. G. J.
Bolten, who has been mentioned on a previous page) under the name of
Pimelodus stegelichii Muller & Troschel, which is, significantly, a synonym
of Rhamdia quelen. Further, that they died in the aquarium of “Artis” on
20 June 1929. Although the registers of the “Artis” aquarium are incomplete,
so that the date on which the specimens were received cannot now be ascer¬
tained, it may perhaps be assumed that they were presented by Dr. Bolten on
his return from Suriname in 1910, and lived to a ripe old age in the aquarium,
which would account for their unusually large size.

In view of the evidence that the two specimens just described are close to
the maximum size the species can attain, I was surprised to see that McConnell
(1967: 65) has recorded specimens of over 12 lb, caught when trawling in
the Rupununi River, and asked for more information. From the reply
received from Mrs. McConnell, I quote: “I ... write to you to confirm your
doubts about identification of “12 lb” Rhamdia quelen. I agree that this

1) Lack of material has prevented me from investigating the validity of R. sapo as
such; I consider it a distinct possibility that it is a synonym of R. quelen.

l6o ZOOLOGISCHE VERHANDELINGEN 132 ( 1974)

does seem much too large, and cannot at this stage see how this crept in
(possibly it was an undetected typing or proof error)”.

Specimen ZMA no. 102234 was recorded by Hoedeman (1961) under the
name of Pinirampus pinirampu. Notwithstanding its somewhat poor and
shrivelled condition, this fish is clearly a perfectly normal specimen of Rham¬
dia quelen .

I am of course aware that, as a subspecies R. q. urichi is here recognized,
correct nomenclatural procedure requires the use of a trinomial for Rham-
dia quelen . The reason why this has not been done is that I have dealt with
the species as a whole, and want to leave open the possibility of geographical
variation within its extensive range. The application of a trinomial would, I
believe, give a false impression of exactness and uniformity.

Rhamdia quelen urichi (Norman)

Caecorhamdia urichi Norman, 1926, Ann. Mag. Nat. Hist., (9) 18: 325, fig. 1 —
Guacharo Cave, Trinidad.

Caecorhamdia urichi ; Borodin, 1927, Amen. Mus. Novit., 263: 2, 3 fig. 13, 5 (Guacharo
Cave, Trinidad); Reichel, 1927, Rev. Suisse Zool., 34: 386, 390 (une grotte de Tile de la
Trinite); Guppy, 1934, Trop. Agricult, n: 117 (Trinidad); Hubbs, 1938, Carnegie Inst.
Wash. Publ., 491: 267 (a Trinidad cave); Gosline, 1941, Stanford Ichth. Bull., 2: 84
(Trinidad); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 41 (Trinidad);
Thines 1955, nn. Soc. Roy. Zool. Belg., 86: 40, 61 fig., 10 (Finterieur de la grotte de
Guacharo (Trinite)); Boeseman, i960, Studies Fauna Curasao, 10: 103 (Trinidad);
Vandel, 1964, Biospeologie: 264 (Cueva del Guacharo, dans File de Trinidad); Thines,
1969, L’Evol. Regr. Poiss. Cavernic. Abyss.: 28, 106 (la grotte de Guacharo (Trinite)).

Caecohamdia urichi ; Wolf, 1934-1937, Anim. Cavern. Cat., 2: 486 (Trinidad: Gua¬
charo Cave); Wolf, 1934-1938, Anim. Cavern. Cat., 3: 697 (Trinidad: Guacharo Cave).

Caecorhamdella urichi ; Hubbs, 1938, Carnegie Inst. Wash. Publ., 491: 264 (Trinidad).

Material. — One specimen, 16 April 1966, Guacharo Cave, Oropuche River, Trinidad
(RMNH no. 26733), standard length 135 mm.

Characters. — Except for the degenerated eyes, described in more detail
below, identical with Rhamdia quelen.

The left orbit is filled with connective tissue, and it is entirey closed by
skin; externally it is merely indicated by a fold of the skin. The right orbit,
however, shows a small hole, less than a millimetre across, in the skin which
covers it. This hole is relatively deep, it appears to be the last vestige of an
open orbit. There is no trace of an eye. Histological studies of the optic
region might yield interesting results, but must await the collecting of more
material.

When this specimen was fresh, I noted that it was “normally pigmented”;
in preservative it looks somewhat paler than two specimens of R. quelen we
captured at the same time and place. Norman (1926; 325) described his
specimens as “uniform pale yellowish brown”, which also describes my pre-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE l6l

served specimen very well. However, some of the individuals we observed did
appear to be white (see Discussion) ; they must have had a strongly reduced
pigmentation.

Norman (1926), as quoted in the Discussion, observed that the mandibular
barbels in the blind fish are somewhat longer than in specimens with developed
eyes. Obviously the idea behind this remark is that in blind individuals tactile
orientation would be more important than in individuals with eyes.

It would be interesting to have this character confirmed on the basis of
more adequate material. Unfortunately I cannot contribute much of a con¬
structive nature as the two specimens of R. quelen captured in the Oropuche
Cave are much larger than the specimen of R. q. urichi. Nevertheless the
measurements of the barbels of these fishes are here given:

R. q. urichi, standard length 135, left maxillary batbel 52, right maxillary
barbel 62, left outer mental barbel 33, right outer mental barbel 40, left inner
mental barbel 25, right inner mental barbel 23 mm. Longest maxillarly barbel
46%, longest outer mental barbel 30%, longest inner mental barbel 19% of
standard length.

R . quelen (1), standard length 200, maxillary barbels ca. 72, right outer
mental barbel 44 (left has been damaged and healed), left inner mental barbel
24, right inner mental barbd 33 mm. Longest maxillary barbel 36%, longest
outer mental barbel 22%, longest inner mental barbel 16^2% of standard
length.

R. quelen (2), standard length 240, left maxillary barbel 43, right maxil¬
lary barbel 74, left outer mental barbel 45, right outer mental barbel 39, inner
mental barbels 30 mm. Longest maxillary barbel 31%, longest outer mental
barbel 19%, inner mental bathels 1234% in standard length.

These measurements illustrate only two points: the great differences that
can exist even between the two sides of one individual, and the relative
decrease in size of the barbels in larger individuals. In view of the considerable
individual variation, very large series would be needed to confirm the differ¬
ence suggested by Norman; at present it is no more than speculation. Ob¬
viously there is no point in taking these measurements from specimens col¬
lected in Suriname or elsewhere outside Trinidad, as the means might vary
geographically. There is the added difficulty that barbels can break and
regenerate, and on the whole I consider it unlikely that there is anything in
this character.

Distribution. — Confined to Guacharo Cave or Oropuche Cave, Trinidad.

Discussion. — The genus Caecorhamdia was introduced for an eyeless
species from Trinidad, with the comment; “Apart from the absence of eyes,
this fish appears almost identical with Rhamdia queleni, Quoy and Gaimard ...

11

162 ZOOLOGISCHE VERHANDELINGEN I32 ( I974)

When species of equal size are compared, however, the mandibulary barbels
appear to be somewhat longer in the blind fish ... In order to draw atten¬
tion to this interesting fish, and following convention, I have erected a new
genus for its reception. I realize, however, that this procedure is somewhat
unsatisfactory, for, were the eyes developed, the specimens from the Guacharo
Cave would probably be identified with the species Rhamdia queleni”.

Notwithstanding this somewhat conditional introduction, the genus has
continued to be recognized in subsequent literature (Gosline, 1941: 84, 1945:
40; Thines, 1955: 61; Boeseman, i960: 103) *). The last-mentioned author
pointed out that the species C. urichi remained known from the original three
specimens only, and that is probably the main reason why subsequent to its
description so few authors have discussed it.

As cave fauna in general has always intrigued me, and as I was eager to
learn more of Caecorhamdia urichi , I spent, on my way home from Suriname,
a few days on Trinidad, where a visit to the Oropuche Cave was arranged
by Dr. V. C. Quesnel. Our party consisted of Dr. V. C. Quesnel and his two
sons, Dr. J. Price, my wife and I. As the cave is on private property (the
Leotaud Estate), we had to obtain a permit to visit it. With this we went
to the house of the overseer, who showed us the way. We walked upstream
along the Oropuche River (here a creek of a few metres wide) for about ten
minutes, until we reached a high mountain wall, from which the river flowed
through a comparatively narrow but high slit or tunnel. This we entered,
wading through the water. After about twenty metres the slit widened out to
a large chamber, where we were greeted by the loud and eery cries of the
Guacharo’s, Steatornis caripensis, from which the cave takes its name.
Through a narrower part we reached a second, smaller, chamber, also in¬
habited by Guacharo’s and finally, through a tunnel in which the ceiling came
down so low that we had to bend to go through, to a third, much smaller
chamber, with a length of no more than twelve to fifteen metres, at the end of
which the wall reached to below water level. Over a bed of pale sand the water
of the Oropuche, no more than two metres wide at this place, welled up from
under the rocks. I estimated that at this place we were about 150 m inside
the mountain.

In this deepest part of the cave there were no birds or bats, but by the light
of our torches we could see several individuals of Rhamdia swimming about.
It was interesting to see that some of the fishes observed appeared to be
white, whereas others were normally pigmented. Contrary to expectation we
found that the white fishes, which we presumed to be eyeless, were by no

1) Hubbs’ (1938: 264) use of the name Caecorhamdella urichi is clearly only a lapsus,
as on another page he uses the name Caecorhamdia urichi .

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

163

means easy to catch. Our total catch during an hour’s work consisted of two
perfectly normal specimens of R. quelen, and only the one individual listed
above which lacks eyes, but is normally pigmented.

Athough the material collected is admittedly insufficient, our impression
was definitely that in the cave the whole scala of intermediates, varying
from normal R. quelen to unpigmented and eyeless individuals (“pure”
Caecorhamdia urichi) was present. In addition. Dr. Price informed us that
at one time there had been a series of specimens, showing this whole range
of variation, in the Zoology Department of the University. During our visit
he searched in vain for this material, but even without it the observations
mentioned and the specimen captured make it clear that according to modem
principles of systematic zoology, Caecorhamdia urichi is neither a valid genus,
nor a valid species, but must be regarded at most as a subspecies freely
hybridizing with normal R. quelen, or perhaps as not even that. In the next
paragraph arguments will be given for regarding it as a valid subspecies,
rather than as a local aberration.

Fig. 34. Section through the Oropuche Cave, Trinidad. This is a rough sketch only,
to assist in obtaining a general idea of the situation; it does not pretend to any degree

of accuracy.

My reasoning in ascribing subspecific status to these fishes is as follows.
In the deepest part of the cave, and only there, R . q. urichi appeared to be
not uncommon, probably more common than R. quelen (the latter would be
less visible as they are always fully pigmented and better able to conceal
themselves). The incidence of blind fishes appears definitely too high to
regard them as accidental freaks of the normal population. On the other hand
the apparent, although at present still insufficiently documented, occurrence
of intermediates makes it clear that R. q. urichi cannot be considered a species.
Per definition, a subspecies must be to a certain extent geographically
separated from other subspecies of the same species. In the cave visited this
condition does not, at first sight, appear to be fulfilled, for that the handful

164

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

of individuals of uric hi inhabiting the deepest part of the cave would be able
to maintain their characters against a presumably frequent immigration of
normal individuals, is out of the question. Therefore I consider it likely that
deeper in the mountain, and inaccessible to man, there is a considerable body
of water, in which uric hi has originally developed, and that the accessible part
of the cave, the upper course of the Oropuche River, is only the outlet of this
large body of water. This means that contact between the eyeless and the
normal populations would remain restricted to the comparatively narrow
outlet, where we found them mixing.

The opinion here expressed finds some support in the large flow of water
appearing from under the rock in the deepest part of the cave. In this con¬
nection it is relevant to record that about two years before our visit two
young men had made an attempt, with diving equipment, to follow the Oro¬
puche upstream below the rock. Working their way down, they became stuck
and drowned. One of the bodies was salvaged by a professional diver, the
other was never found but some six months after the accident a human skull
was picked up downstream outside the cave. This tragedy had caused an
understandable reluctance to resume underwater exploration in the cave.

It may be noted that the situation as we found it is somewhat different
from that described by Norman (1926), according to whom the type material
was taken: "... from a pool in the interior of the Guacharo Cave, Trinidad ...
this pool is always in complete darkness; very occasionally, in times of heavy
rains, the pool becomes connected with a rivulet running out of the cave”.
During our visit we did not find any pools not connected with the rivulet, but
it is possible that periodically there is less water and the rivulet ceases to run.
Thus the free contact between populations in the cave and outside may be
periodically broken. Carriker (1931), in his picturesque description of the
cave, recorded a considerable flow of water in June.

The relation between Rhamdia quelen urichi and Rhamdia quelen has an
evident parallel in that between Pimelodella kronei and Pimelodella transitoria
in south-eastern Brazil as described by Pavan (1946). There also, the eyeless
form was originally described as a separate genus, subsequently reduced to
a “variety” by Haseman (1911), but finally treated as a separate species by
Pavan. In the case of Pimelodella , Pavan provided evidence (but inconclusive)
that hybridisation between P. kronei and P. transitoria does not normally
occur.

An exhaustive discussion of the problems relating to cave faunae falls outside
the scope of this paper; for more information on the subject reference may
be made to Thines’ (1969) excellent work, but I would like to give my opinion
on the development of cave fishes in three points.

MEES, AUCHENIPTERIDAE AND PIMELODTDAE l6.S

i a Caves will for certain species form an attractive habitat, mainly because
of the absence of predation, but also because of diminished interspecific com¬
petition. Several authors have stressed the need for “preadaptation”: it is
obvious that generally only those species can carry out a successful initial
colonisation of cave habitats, which are (for finding food and other primary
functions) to a large extent independent of sight. 2° The eyes, as well as the
pigment, or rather the melanophores, are highly evolved systems which can
only be maintained by a rigid selection pressure for their retention. Where
this selection pressure falls away, a comparatively rapid degeneration, going
initially through a stage of great variability, is predictable. In a previous paper
I referred to this process as “a loss of characters as a consequence of the
absence of selective pressure for their retention” (Mees, 1962: 32). I con¬
sider therefore that the degeneration of eyes and melanophores is an inevitable
consequence of life in darkness where these organs are not needed. 3 0 In
the case of the eyes there may well be selection pressure for their loss, as in
total darkness eyes would probably be easily injured; once the eyes have
degenerated and the sockets closed by strong skin, a possible source of injury
and infection has been eliminated. Following this line of reasoning, it is clear
why the degeneration of the eyes precedes the loss of melanin (admittedly
Hubbs, 1936: 168, 182, recorded that Rhamdia guatemalensis decolor from
caves near Motul in Yucatan, shows depigmentation without reduction in
eye-size). Indeed, most blind fishes possess some pigment; this is also as
could be expected: I have on purpose distinguished between the melano¬
phores, which are specialized structures, and the pigment itself, the produc¬
tion of which is only one of the functions of the melanophores. Therefore
it is not surprising that certain eyeless species are still able to form pigment
(cf. Thines, 1969: 234), but appear unable to manipulate it. Actually, as
Thines (1969: 235) has suggested, the absence of pigment may also have
selective value in cave fishes.

Rhamdia foina (Muller & Troschel) (fig. 33)

Pimelodns foina Muller & Troschel, 1848, in R. Schomburgk, Reisen in Britisch-
Guiana, 3: 628 — besonders an den felsenreichen Stellen des Takutu.

Rhamdia holomelas nipununi Fowler, 1914, Proc. Acad. Nat. Sci. Philad., 66: 258,
fig. 11 — Rupununi River, British Guiana.

Pimelodus foina ; Muller & Troschel, 1849, Horae Ichth., 3: 5 (Guiana); Gunther,
1864, Cat. Fish. Brit. Mus., 5: 130 (Guiana).

Rhamdia foina ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2) 1: 126
(name only); Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1 : 126
(Guiana); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 28 (Takutu,
Guiana); Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 386 (Guiana).

Rhamdella foina ; Eigenmann, 1912, Mem. Carnegie Mus., 5: 167 (Warraputa Catar-

i66

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

act) ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool, 33: 34 (Guiana); Alexander,
1966, Proc. Zool. Soc. Lond., 148: 90 (British Guiana).

Rhamdia holomelas rupununi ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool.,
33 ■ 36 (rio Rupununi).

Material. — One specimen, 1842, Takutu (R. Schomburgk, ZMB no. 3039), standard
length 163 mm, holotype of the species. One specimen, 1908, Warraputa Cataract, British
Guiana (C. H. Eigenmann, CAS no. 13468, ex IUM no. 12007), standard length no mm.

Characters. —- D 1 . 6 , A io-ii (iv.6 and iv.7), P 1.8 and I.9, V 6 (1.5),
C i.14.1 and i.i5.i, excluding rudiments, branchiostegals 6, gill-rakers 17-19
(4 + 1 + 12 and 4 + 1 + 14), well-developed, long and slender. Head
4.04.2 in standard length, greatest width of body, between the cleithra,
5.2-5.35 times in standard length, depth of head at the same place 0.6-0.7
times its width, depth of body 6.S-6.7 times in standard length, predorsal
length 2.9-3.0 times in standard 'length, adipose base 34-37 times in standard
length, eyes 1.7-2.0 times in snout and 44-4.8 times in head, the maxillary
barbels reach to below the dorsal fin, dorsal and pectoral spines well-
developed, the former (broken in the type) 2.8, the latter 2.3 and 2.3 times
in predorsal length. This species differs from R. quelen by its conspicuously
larger eyes, more depressed snout, rather stronger spines, shorter adipose fin
and higher number of gill-rakers. For shape of fontanel and occipital
process, see fig. 33.

Distribution. — On present evidence confined to British Guiana, where
known from the Takutu and Rupununi Rivers, and the Warraputa Cataract,
Essequibo River. As the Takutu River belongs to the Amazon basin, a much
wider distribution is likely.

Discussion. — The reason I studied this species, is that Eigenmann (1912)
had placed it in Rhamdella, hence in the same genus as Imparfinis minutus.
As at the same time it seemed unlikely that I. minutus, which is distributed
throughout Suriname and has an enormous range in eastern South America,
would not occur in British Guiana, I suspected that R. foina might be the
same as /. minutus, over which it has ample priority. On examination I
found, however, that R . foina is an entirely different species, which appears
to be well-placed in the genus Rhamdia. If, however, a separate genus Rham¬
della is recognized, exclusively based on the shape of the fontanel, R . foina
would probably belong to it.

The species described by Fowler (1914) under the name of Rhamdia
holomelas rupununi had puzzled me considerably. R. holomelas has been
generally regarded as a synonym of R. quelen , and R. h. rupununi could have
been expected to belong in the same relationship. Nevertheless, Fowler’s
description and illustration made me very reluctant to unite R . h. rupununi
with R. quelen, as there were too many obvious discrepancies. This problem

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

167

was solved when I received specimens of R. foina, and found that Fowler’s
description fits them perfectly.

Genus Imparfinis Eigenmann & Norris

Imparfinis Eigenmann & Norris, 1900, Rev. Mus. Paul., 4: 351 — type by monotypy,
Imparfinis piperatus Eigenmann & Norris.

Rhamdioglanis R. von Ihering, 1907, Not. Prel. Mus. Paul., 1 (1) : 16 — type by
monotypy, Rhamdioglanis frenatus R. von Ihering.

Nannorhamdia Regan, 1913, Ann. Mag. Nat. Hist., (8) 12: 467 — type by monotypy,
Nannorhamdia spurrellii Regan.

Generic diagnosis. — Resembling Rhamctia, but posterior border of skull
as if cut off straight, with but a rudimentary postoccipital process; fontanel
a long slit, reaching to the occiput; dorsal and pectoral spines present but
very inconspicuous, continued as soft rays, the soft part usually longer than
the bony part; maxillary barbels short to moderate in length, varying from
scarcely reaching pectoral base, to reaching to the end of the anal fin; eye-rim
free 1 ); no pectoral pore; adipose fin comparatively short.

Distribution. — Throughout tropical South America, east and west of the
Andes, and Central America, reaching Costa Rica (Pacific drainage).

Discussion. — The first attempt to separate fishes belonging here generically
from Rhamdia , was made by Eigenmann & Eigenmann (1888), who intro¬
duced a subgenus Rhamdella (raised to generic rank in a postscript to the
same paper), in which they included several species previously referred to
Rhamdia , one of these being Rhamdia minuta Liitken.

In the original description of Rhamdella , the Eigenmanns (1888) mentioned
only a single character to differentiate it from Rhamdia ; “fontanel continued
to the occiput, a bridge across it behind the orbit”. In their next publication
the Eigenmanns (1890: 137) expanded the description as follows: “Fontanel
continued to the occipital process, a bridge behind the eyes. Dorsal and
pectoral fins with strong spines. Ventrals placed below or behind the dorsal.
A free orbital margin”. Although Rhamdia minuta does certainly not have
strong dorsal and pectoral spines, Eigenmann & Eigenmann (1890) retained
this species in Rhamdella without comment. There is even a discrepancy
between the diagnosis of Rhamdella and the description of its type-species,
R. eriarcha , where we find: “the first ray of the dorsal scarcely spine-like,
smooth”.

1) In small specimens (less than 40 mm in standard length), the eye-rim is little
developed, especially above the eye; in poorly preserved material, showing the “pop-
eyed” condition, the eye may appear to have no free rim, and has sometimes been
described as such. Well-preserved material, if not too small, rarely gives any trouble.

i68

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

A re-diagnosis of Rhamdella, as well as a renewed study of the species
currently placed in it, was evidently required. To arrive at a satisfactory
classification, three questions had to be answered: (i) are Rhamdella eriarcha
and Rhamdia minuta congeneric; (2) if not, which genus does Rhamdia
minuta belong to; (3) is Rhamdella eriarcha sufficiently distinct from Rham
dia to be retained in an own genus, when (1) has been answered in the nega¬
tive.

Unfortunately, regulations forbade that the type specimen of Rhamdella
eriarcha, which remains an unicum, was sent on loan to me, but Miss Dick
has provided me with a description and measurements from which I consider
it justified to deduce that it is much closer to typical Rhamdia than to Rham¬
dia minuta ; these characters are: well-developed, sturdy pectoral spines; a
weak dorsal spine; a long adipose fin, less than three times in standard length
(actual measurements: standard length 191 mm, adipose fin 69 mm), and a
well-developed postoccipital process. Against this I consider the short maxil¬
lary barbels of Rhamdella eriarcha as less important; the presence of a well-
developed postoccipital process causes the fontanel, which admittedly is an
elongated slit, to be not directly comparable to that of R . minuta.

The conclusion is that, unless the currently accepted generic limits are
broadened and all these forms are united with Rhamdia , R. minuta is not
congeneric with Rhamdella eriarcha . Further, it seems to me that Rhamdella
eriarcha is so close to Rhamdia , that its retention in a separate genus is
hardly justified; this question, however, can better be left open until such
time as a direct comparison between Rhamdella eriarcha and the many species
currently placed in Rhamdia and Rhamdella can be made.

Remains the second question, which genus R. minuta belongs to. The oldest
name that seemed likely was Imparfinis Eigenmann & Norris, as the authors
of that genus described its type-species, /. piperatus, as being very similar
to Rhamdia minuta. There are, however, two discrepancies in the description,
which would preclude the inclusion of R. minuta in the genus Imparfinis;
the first is that Imparfinis was described as having no free eye-rim, the second
that it was supposed to have a patch of teeth on the vomer (a most ususual
character for any species of the Rhamdia-gronp) . The holotype of Imparfinis
piperatus is in the California Academy of Sciences, where Dr. Eschmeyer
has examined it; he informed me as follows: “in the type it would appear
that maybe ventrally the eye-rim is present, but dorsally the eye is more or
less tightly covered by continuous skin, i.e., skin from outside eye continues
on about the same plane over the eye”, and: “I see small teeth in the jaws
but cannot locate any on the vomer or palatines. Someone has removed much
of the skin on the roof of the mouth. If this did not rub off the teeth, then

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

169

no vomerine teeth were present”. The type is a specimen of only 40 mm
total length. In specimens of this size the eye-rim is usually little-developed,
and as Dr. Eschmeyer has also informed me that the type has a rudimentary
postoccipital process, similar to that of R. minuta, I accept this as evidence
that Imparfinis piperatus is closely related to, and certainly congeneric with,
Rhamdia minuta. The last-mentioned species has, therefore, to be known as
Impar finis minutus.

I have not tried to study all members of the genus Impar finis as here
defined, but at least the following species appear to belong to it: Rhamdia
microcephala Liitken, 1874; Rhamdia minuta Liitken, 1874; Impar finis
piperatus Eigenmann & Norris, 1900; Nannorhamdia spurrellii Regan, 1913;
Nannoglanis hoehnei A. de Miranda Ribeiro, 1914; Nannorhamdia nema -
cheir Eigenmann & Fisher, 1916; Nannorhamdia guttatus Pearson, 1924;
Nannorhamdia stictonotus Fowler, 1940; Pimelodella cochahambae Fowler,
1940; Nannorhamdia macrocephala Miles, 1943; Nannorhamdia benedettii
Fernandez & Martin, 1952; Nannorhamdia schubarti Gomes, 1956; and
Nannorhamdia lineata Bussing, 1970.

On the other hand, several species described as Imparfinis, clearly do not
belong to the genus. Imparfinis tenebrosus Schubart (1964) has very small
eyes, evidently without free rim, and in the description no mention is made
of the presence of spines in D and P. Imparfinis bolivianus Pearson (1924)
was described as having small eyes without free rim, and there is no mention
of spines. Both these species are probably referable to the genus Heptapterus ;
they should be known as Heptapterus bolivianus (Pearson) and Heptapterus
tenebrosus (Schubart).

The generic allocation of Rhamdella schultzi P. de Miranda Ribeiro
(1964b) is doubtful. It was described as having the eyes entirely subcutaneous
and very small (2 mm diameter in a specimen of 94 mm standard length), and
therefore can belong neither to Rhamdia-Rhamdella, nor to Imparfinis. It
may be another member of the genus Heptapterus , but illustration and descrip¬
tion indicate a heavier body than is usual for that genus.

Haseman (1911: 318) stated: if Imparfinis transfasciatus (Ribeiro) and
Rhamdiaglanis frenatus (R. Von Ihering) are synonyms of Imparfinis pipera¬
tus. I have examined the types of all these species”. The specimen described
and figured by A. de Miranda Ribeiro (1911: 260, pi. 44 fig. 1, not pi. 39)
had a total length of 34 cm, and the figure shows a heavy-bodied fish,
entirely different from Rhamdia minuta with which Eigenmann & Norris
(1900) compared Imparfinis piperatus . Although R. transfasciatus has for
sixty years been almost universally regarded as a synonym of I. piperatus,
I doubt that this is correct. Alternately, a mix-up may have occurred in A. de

170

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

Miranda Ribeiro’s illustrations, and the fish captioned Rhamdioglanis trans-
fasciatus may not be that species; indeed the figure would suggest Pseudo-
pimelodus zungaro to me, were it not that the ventral fins appear to originate
well in front of the dorsal origin, a character not found in either Pseudo-
pimelodus or the supposed specimen of R. transfasciatus discussed in the
next paragraph. I have to leave the solution of these problems to a future
reviser.

The specimen listed under the name Rhamdioglanis transfasciatus by van
der Stigchel (1946: 47) is very similar to /. minutus but has smaller eyes and
a more elongated body; the markings are as in the more strongly marked
specimens of L minutus , it is longer than any specimen of that species I
examined, and rather more slender (which is significant in connection with
the remarks made above about A. de Miranda Ribeiro’s specimen). For
comparison with I. minutus I give here some particulars on the specimen: it
is undated, from Joinville, Est. Sta. Catarina, ex Mus. Vienna (1909), det
Steindachner, ZMA no. 110696. Standard length 107 mm, predorsal length
3.0, head 4.7, body width between cleithra 6.05, depth of body 7.5, distance
from snout to origin of ventrals 2.5, adipose fin 3.75 times in standard length,
maxillary barbels reaching to just beyond pectoral base, eye-rim only just
free, eye 2.2 times in snout, 5.5 times in head, 1.0 times in bony interorbital,
1.7 times in entire interorbital. D 1 . 6 , A 12 (iv. 8 j 4 ), P I.9, V i.5, ventral
origin below third or fourth dorsal ray, branchiostegals 8, gill-rakers n
(2 + 1 + 8).

Imparfinis minutus (Liitken) (fig. 33)

Rhamdia minuta Liitken, 1874, Oversigt K. Danske Vidensk. Selsk. Forh.: 35 — no
locality = Rio das Velhas.

Imparfinis hasemani Steindachner, 1915, Denkschr. Akad. Wiss. Wien, Mathem.-
Naturw. Kl., 93: 59, fig. 1-3 — aus dem Rio Surumu bei Serra do Mello, dem Rio branco
bei Bern Querer und dem Rio Tapajos bei Santarem.

Cetopsorhamdia pijpersi Hoedeman, 1961, Bull. Aquatic Biol., 2: 132 — Sipaliwini.
Rhamdia minuta\ Liitken, 1875, K. Danske Vidensk. Selsk. Skr., (5) 12: 179, pi. 3
fig. 6 (Rio das Velhas); Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci., (2) 1:

131 (Rio Janeiro, Macacos); Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3:
387 (Rio de Janeiro; Macacos); A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de
Janeiro, 16: 264, fig. 108 (Rio das Velhas, Rio de Janeiro — Ribeirao das Lages?,
Macacos),

Rhamdella minuta ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1:
142 (Rio Janeiro; Macacos); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus.,
14: 28 (Macacos; Rio das Velhas; Rio de Janeiro); Fowler, 1940, Proc. Acad. Nat Sci.
Philad., 92: 80 (Rio das Velhas); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool.,
33 ’ 35 (Rio de Janeiro, Macacos); Fowler, 1951, Arq. Zool. S. Paulo, 6: 565 (Rio Sao
Francisco, Rio de Janeiro, Macacos, Sao Paulo).

Cetopsorhamdia hasemani ; Gosline, 1941, Stanford Ichth. Bull., 2: 87 (Rio Branco

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

I7I

basin, State of Amazonas, Brazil) ; Gosline, 1945, Bob Mus. Nac. Rio de Janeiro, Zool.,
33: 31 (bacia do Rio Branco); Hoedeman, 1961, Bull. Aquatic Biol., 2: 132 (the Rio
Branco system).

Imparfinis hasemani; Fowler, 1951, Arq. Zool. S. Paulo, 6: 527 (Amazonas).

Material. — 72 specimens from Suriname as follows. One specimen, 6 February 1961,
Sipaliwini (H. P. Pijpers, ZMA no. 102230), standard length 91 mm, holotype of
Cetopsorhamdia pijpersi. Three specimens, 6 March 1964, rapids in the Gran Rio not far
from its mouth (Boeseman, RMNH no. 25768), standard length 67, 82 and 89 mm. One
specimen, 21 July 1964, from a tributary of the Gran Kreek, about twelve kilometres
upstream of the mouth of the latter (Boeseman, RMNH no. 25769), standard length
60 mm. Three specimens. 31 July 1964, rapids in the middle course of the Gran Kreek
(Boeseman, RMNH no. 25770), standard length 53^, 60 and 70 mm. 15 specimens,
13 August 1965, Mamadam, Suriname Rivier (Mees, RMNH no. 25771), standard length
32-71 mm. Three specimens, 13 September 1965, Avanavero Vallen, Kabalebo Rivier
(Mees, RMNH no. 25772), standard length 49^2, 50, 54 mm. Two specimens, 17 Novem¬
ber 1965, Tapanahoni Rivier, about two kilometres downstream of its confluence with
the Paloemeu (Mees, RMNH no. 25773), standard length 49, 72 mm. Five specimens,
20 October 1966, Gran Kreek, 63 km S. of Affobakka (Nijssen, ZMA no. 106250),
standard length 48, 51, 52, 53, 54 mm. 21 specimens, 5 April 1967, Stondansie Vallen,
Nickerie Rivier (Nijssen, ZMA no. 106249), standard length 43-80 mm. Six specimens,
5 April 1967, tributary of Nickerie Rivier, ca. 12 km S. of Stondansie (Nijssen, ZMA
no. 106261), standard length 62-68^ mm. 14 specimens, 6 April 1967, Fallawatra Rivier,
5 km S. of Stondansie (Nijssen, ZMA no. 106248), standard length 62-84 mm. Four
specimens, 10 August 1968, Kaysergebergte (M. S. Hoogmoed, RMNH no. 25774),
standard length 21, 22, 22, 22 mm.

Extra-limital material. Two specimens, no date, Rio das Velhas (J. Reinhardt, UZMK
nos. 325, 328), standard length 63, 71 mm, syntypes of the species. One specimen, 1864,
Macacos (W. M. Roberts, MCZ no. 7626), standard length 90 mm. One specimen,
December 1909, Santarem (J. D. Haseman, MV no. 46090), standard length 66 mm,
syntype of Imparfinis hasemani. One specimen, ca. 1909, Bern Querer, Rio Branco
(J. D. Haseman, MV no. 46091), standard length 63 mm, syntype of Imparfinis hasemani.
One specimen, ca. 1909, Bern Querer, Rio Branco (J. D. Haseman, MV no. 46093),
standard length 89 mm, syntype of Imparfinis hasemani. One specimen, 25 March 1968,
Sangadina stream at is confluence with the Rio das Mortes, Mato Grosso (R. H.
McConnell, BM no 1971.7.29: 1), standard length 38 mm. One specimen, 19 April 1968,
stream near Xaventina, Rio das Mortes (R. H. McConnell log 95), standard length
73 mm. One specimen, 1 May 1968, Xaventina Island, Rio das Mortes (R. H. McConnell
log 120), standard length 48 mm. Two specimens, same data, standard length 44, 44 mm.
One specimen, no date, Rio Negro (E. Goeldi, MCZ no. 27274), standard length 63 mm.

Diagnosi's. — A typical member of the genus. The bony part of the head cuts
off straight at the nape, with but a very short or vestigial occipital process,
and the distance from the tip of the snout to the posterior edge of the occiput
is the same or only a little more than the distance from the occiput to the
origin of the dorsal fin. Further characterized by its very weak spines in
dorsal and pectoral fins. Adipose fin short, barbels short, maxillary barbels
not reaching to the tips of the pectoral fins.

Description. — D 1 . 6 , A 9-11 (iii.6-iv.7>^), P I.9 or 1 .10, V 6 (i.5), C i.i5.i
and rudiments, gill-rakers on outer branchial arch 13 to 16 (1-2 + 1 +
11-13), branchiostegails 7 or 8, vertebrae 38.

172

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

A fairly small (21-91 mm in standard length) and slender species. The
body is more or less cylindrical, its anterior part on an average slightly wider
than deep, its posterior part compressed. Width of body (between the pec¬
torals) 5.6-7.0 in standard length, depth of body 5-8-8.2 in standard length,
distance from the tip of snout to origin of dorsal fin 2.75-3.0 in standard
length.

Head, measured from tip of snout to posterior border of operculum,
4-45“5-9 m standard length, the bony structure only covered by thin skin,
occipital process vestigial, fontanel long and narrow, continued to the occiput.
Eyes large, 3.8-4*5 in head, with a free rim; width of bony interorbital
about three-quarters of eye diameter. Mouth inferior; upper and lower jaw
each with a band of teeth; the band of teeth in the upper jaw not clearly
interrupted at the median, and without backward extensions; the band in the
lower jaw narrower, distinctly interrupted at the median, no teeth on vomer
or palatines. The posterior pair of nostrils slightly less than half an eye's
diameter in front of the eyes, and a distance apart which is similar to the
width of the bony interorbital, roundish, without rim or flap; the anterior
pair of nostrils is placed behind the upper lip, at the end of small thin tubes;
the shortest lines connecting the four nostrils, viewed from above, make
almost a square. There are three pairs of barbels; the maxillary pair of
barbels is implanted laterally from the anterior nostrils, and reaches to the
posterior half of the pectorals (usually about two-thirds or three-quarters
along the pectorals when these are pressed against the body), and never as
far as the vertical through the origin of the dorsal fin; the outer pair of
mental barbels reaches to the origin of the pectorals or very little beyond; the
inner pair of mental barbels is about two-thirds the length of the outer pair.
Gill-rakers well-developed, slender and fairly long, the longest about half an
eye's diameter; 13 to 16 on the outer branchial arch, usually divided as
follows: 1 or 2+ 1 + utoi3 (in one specimen I counted only twelve gill-
rakers, 2+1+9, t> ut as ^e gills were a bit slimy and not very well-pre¬
served, one or two might have been missing).

Lateral line complete, almost straight, its anterior part slightly arched.

Dorsal fin with a very weak spine, continued as a simple ray, the soft part
being of the same length, or a little longer, than the spine, and six divided rays,
the first two rays (one simple and one divided) of equal length, the following
progressively shorter, and the last ray half or just over half the length of the
first ray; length of base of dorsal equal to snout and eye; longest rays of
dorsal equal to head.

Anal fin with three or four simple and six or seven divided rays, the last
one of which is usually split right down to its base; it is inserted well back-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

173

wards, the distance from its origin to the tip of the snout being more than
twice the distance from its origin to the base of the caudal fin; the length
of its base is about the same as the length of the dorsal base, or a trifle less;
outline of fin rounded, longest rays equal to snout and eye.

Pectorals with one spine and nine or ten divided rays; spine weak, with
along its anterior margin about 15 antrorse hooks, which become smaller
distally; length of spine equal to distance from tip of snout to anterior margin
of pupil; the spine is continued as a simple ray, and this part is about three-
fifths to three-fourths of the length of the spine, or equals it, whereas both
parts together are equal in length to the distance from tip of snout to occiput,
hence almost as long as the head; the following rays are successively shorter,
giving the fins a pointed and even slightly falcate appearance; pressed against
the body the pectorals cover about three-fourths of the distance from their
implantation to the ventrals.

Ventrals with one simple and five divided rays, obliquely rounded with the
second and third rays longest; the ventrals are only a little shorter than the
pectorals, and pressed against the body they cover two-thirds of the distance
from their origin to the origin of the anal fin, or slightly less.

Caudal fin long and deeply forked, with acute lobes, usually with 15 divided
rays, of which seven in the upper lobe, seven in the lower lobe, and one
exactly symmetrical ray in the middle. Above and below, the fin has one
simple ray which is of the same length as the longest divided ray, outside
that one a simple ray of about half its length, and subsequently several pro¬
gressively smaller rays.

Adipose fin well-developed, but not very long, its base 4.g-6.6 (usu¬
ally 5.3-5.8) in standard length, rounded, or bluntly triangular, its height
about half the depth of the body below it, or a little less. Its origin is over the
origin of the anal fin or perhaps very slightly in advance of it.

Colour in life light grey-yellow, pale below. In a preserved condition the
colour is light greyish brown; the largest specimens have on the nape, at the
origin of the dorsal fin, and very faintly farther backwards, concentrations
of blackish pigment which can be seen as very faint dark cross-bars. Large
specimens have also pigment concentrations on the lateral line and on the
middle of the back between dorsal and adipose fin. The smallest specimens
have a narrow dark stripe from the middle of each eye over the snout to the
upper lip; in larger specimens this stripe becomes indistinct.

Habitat. — In Suriname this species is on present evidence confined to
the rivers and large creeks, on places with a strong current. The specimens
from the Mamadam were taken in growths of Podostemaceae on rocks in a
strong current; they were associated in this place with Heptapterus longior.

174

ZOOLOGISCHE VERHANDELINGEN 13 2 (1974)

The specimens from the Tapanahoni and the Avanavero Vallen were similarly
taken in rapids. Ecologically the species appears to agree with Heptapterus
longior and Heptapterus tapanahoniensis, but is less common.

Distribution. — The species is probably distributed throughout Suriname,
as indicated by the fact that it was found in a number of widely separated
localities, in four different river systems. It has not yet been recorded from
French and British Guiana, but almost certainly occurs there as it is known
from the basins of both boundary rivers, the Corantijn in the west, and the
Marowijne in the east.

In Brazil known from the Amazon and several of its major tributaries,
the Rio das Velhas (Rio Sao Francisco basin); Rio de Janeiro. Records from
the State of Sao Paulo are doubtful (see Discussion).

The species is unknown from the comparatively well-studied Essequibo
River, or from anywhere farther west, so that the Corantijn may constitute its
limit of distribution in northern South America. Venezuela and Colombia
have different species (/. nemacheir, /. benedettii), easily distinguished by
their longer maxillary barbels (reaching to well beyond ventral base), and
longer adipose fin.

Discussion. — It is difficult to understand why Hoedeman (1961) placed
his specimen in the genus Cetopsorhamdia, with the statement that it: “agrees
with the generic diagnosis for the genus Cetopsorhamdia as defined by Eigen-
mann & Fisher (1916) and criticized by Gosline (1941) and Schultz (1944)”.
The genus was expressly diagnosed as having no free orbital rim, a character
usually given much weight in the classification of the Pimelodidae. Though
the value of this single character has possibly been overestimated, it should
certainly not be altogether ignored.

The description given here was based on specimens from Suriname; two
syntypes of the species (from the Rio das Velhas) differ in that the base
of the adipose fin is slightly longer, 4.7 and 4.7 times in standard length, and
the caudal peduncle more slender (about four-fifths of the depth of Suriname
specimens of similar size); one of these specimens has D 1 .5 instead of the
usual 1 . 6 . A specimen from Macacos has the base of the adipose fin 4.9 times
in standard length; specimens from the Rio das Mortes have it 5.35-5.6 in
standard length, and three syntypes of I. hasemani 4.7, 4.7 and 5.0. A speci¬
men from the Rio Negro (MCZ no. 27274) has the maxillary barbels very
long, reaching to beyond the ventral base, gill-rakers 1 + 1 + 8, and adipose
fin 4.5 times in standard length. It is also possible that Amazonian specimens
are slightly more strongly pigmented than specimens from Suriname, but it
is my opinion that none of the mentioned differences is enough to justify
expression in nomenclature, so that I refer them all to I. minutus.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

175

The specimens from the Rio das Velhas which Fisher (1917: 409) referred
“without question” to this species, under the name of Imparfinis minuta, were
described as having the maxillary barbels reaching to near the tips of the
ventrals. In the two syntypes of R. minuta, one had the maxillary barbels
reaching to near the tips of the pectorals, the other to the tips of the pectorals,
and no specimen examined by me has them to the tips of the ventrals.

Judging by its description and figure, Pimelodella cochabambae Fowler
(1940b: 80) is extremely close to I. minutus, from which it appears to differ
only in the more forward position of the dorsal fin (in the figure I measure
a predorsal length of about 3.4 times in standard length, as against no more
than 3.0 in I. minutus), the slightly longer maxillary barbels, somewhat
smaller eyes (about 5 times in head), and perhaps by having larger fins,
including a rather long adipose base. P. cochabambae is known from a single
small specimen only (Fowler gives a length, evidently total length, of 79 mm,
but measuring from the figure the standard length would be no more than
50 mm), and additional material might well show that it is only a subspecies
of L minutus and perhaps not even that.

Fowler (l.c.) described P. cochabambae as: intermediate between Nan -

norhamdia and Rhamdella . It largely agrees with the former in the short
supraoccipital and humeral extensions, but differs in the distinct dorsal and
pectoral spine. Apparently closest to Rhamdella minuta (Liitken) On
the basis of this opinion of its affinities, it is not obvious to me why Fowler
should have described P . cochabambae in the genus Pimelodella.

Imparfinis spurrellii (Regan)

Nannorhamdia spurrellii Regan, 1913, Ann. Mag. Nat. Hist., (8) 12: 467 — R. Condoto.

Nannorhamdia spurrelli; Eigenmann, 1922, Mem. Carnegie Mus., 9: 40 (Istmina, San
Juan Basin).

Nannorhamdia spurrellii; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 40
(bacia do San Juan, Colombia) ; Bussing, 1970, Contrib. Sci. Los Angeles County Mus.,
196: 2-7 (Rio San Juan drainage, Colombia).

Material. — One specimen, undated, Rio Condoto (Spurrell, BM no. 1913.10.1 : 41),
standard length 62 mm, holotype of Nannorhamdia spurrellii.

Characters. — Very similar to I. minutus, but eyes smaller, ca. 5.5 in head,
2.0 in snout, 1.5 in entire interorbital and 1.0 in bony interorbital.

A few additional counts and measurements are given here for comparison
with I. minutus. D. 1 . 6 , A 10 (iii.7), P 1 - 9 , V 6 (i.5), C 16 (only counting
well-developed rays; as both tips were broken and missing, it was not possible
to distinguish between simple and divided rays), gill-rakers on outer bran-

176

ZOOLOGISCHE VERHANDELINGEN 13 2 (I974)

chial arch (2 + 1 + 10 or perhaps more correct 3 + 10) x ), branchiostegals
9. Width of body 5.75 times in standard length; depth of body 6.0 in standard
length, distance from tip of snout to origin of dorsal fin 3.0 in standard
length; head 4.5 in standard length, predorsal length 3.0 in standard length,
adipose base 5.0 in standard length; maxillary barbels just reaching to a
vertical through origin of D, outer mental barbels reaching to a little beyond
origin of P, inner mental barbels about two-thirds length of outer mental
barbels; origin of V opposite third divided ray of D, length of pectoral spine
about 0.4 times length of spine and ray together; caudal base slender as in
the types of I . minutus ; a dark lateral band exactly as in specimens of I.
minutus from Suriname; no clear indication of cross-bands. All these charac¬
ters fall in the range of variation of L minutus, except that the pectoral spines
tend to be proportionally a little shorter, and that the number of branchio¬
stegals is high. Series would be required to evaluate these points, and I am
disinclined to attach much value to them.

Distribution. — Known from two specimens, the type from the Rio Con-
doto, and a specimen from Istmina, very near Condota, and both San Juan
basin.

Discussion. — This species is very close to /. minutus and one would be
inclined to give it no more than subspecific status.

Imparfinis hoehnei (A. de Miranda Ribeiro)

Nannoglanis hoehnei A. de Miranda Ribeiro, 1914, Comm. Linh. Telegr. Estr. Matto-
Grosso ao Amazonas, 15, Annexo 5: 3, pi. 1 fig. 1 — Rio Taquary, Matto Grosso.

Imparfinis hoehnei ; Gosline, 1941, Stanford Ichth. Bull., 2: 87 (Rio Taquary, State
of Mato Grosso, Brazil); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 31
(Mato Grosso); Fowler, 1931, Arq. Zool. S. Paulo, 6: 527 (Rio Paraguai).

Nannoglanis hoehnei ; P. de Miranda Ribeiro, 1953, Arq. Mus. Nac. Rio de Janeiro,
42: 403 (Rio Taquari, Mato-Grosso); Schubart, 1964, Bol. Mus. Nac. Rio de Janeiro,
Zool., 244: 10 (Rio Taquari, Mato Grosso).

Material. — None.

Characters. — A. de Miranda Ribeiro’s (1914) plate shows a fish very
similar to /. minutus, but easily distinguished by longer head and anterior
part of the body; from the figure I measured: head 4.2 times in standard
length, predorsal length 2.5 times in standard length.

Distribution. — Only known from the two syntypes from the Rio Taquari,
Matto-Grosso, Paraguay-La Plata drainage.

1) Bussing (1970) gives 2+10 for this same specimen, counted for him by Mr.
Wheeler. The first raker on the epibranchial is somewhat rudimentary, it has been
included by me, but has apparently not been counted by Mr. Wheeler.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

177

Genus Heptapterus Bleeker

Heptapterus Bleeker, 1858, Visschen Ind. Arch., 1 : 197 — type by monotypy, Pimelodus
mustelinus Valenciennes.

Pariolms Cope, 1872, Proc. Acad. Nat. Sci. Philad., (1871) : 289 — type by monotypy,
Pariolius armillatus Cope.

Nannoglanis Boulenger, 1887, Proc. Zool. Soc. Lond.: 278 — type by monotypy,
N anno giants fasciatus Boulenger.

Acentronichthys Eigenmann & Eigenmann, 1889, Proc. Calif. Acad. Sci., (2) 2: 28 —
type by original designation, Acentronichthys leptos Eigenmann & Eigenmann.

Chasmocranus Eigenmann, 1912, Mem. Carnegie Mus., 5: 131 (in key), 160 — type
by original designation, Chasmocranus longior Eigenmann.

Imparales Schultz, 1944, J. Wash. Acad. Sci., 34: 93 — type by original designation,
Imparales mariai Schultz.

Medemichthys Dahl, 1961, Noved. Colomb., 1 : 490 — type by original designation,
Medemichthys guayaberensis Dahl.

Phenacorhamdia Dahl, 1961, Noved. Colomb., 1: 504 — type by original designation,
Phenacorhamdia macarenensis Dahl.

Generic diagnosis. — Long and slender species with depressed head,
lacking spines in D and P, and with the upper surface of the head entirely
covered with a layer of connective tissue and thick skin, showing no exposed
bone; barbels short or moderate in length, reaching at most to base of
ventrals, and usually much shorter; eyes small, dorsal in position, without free
orbital rim; teeth in bands in both jaws, with or without backward projec¬
tions laterally, none on vomer and palatines; dorsal fin short, with seven
rays (eight in one species); anal fin varying from 8 to over 30 rays; ventrals
six-rayed, implanted opposite, or a little behind, or a little in advance of
dorsal origin,; adipose fin long, either broadly or narrowly connected with
the caudal fin, or clearly separated from it; caudal fin of any shape. Colour
usually earth brown, sometimes with a few whitish markings or with some
ill-defined blackish cross-bands.

Discussion. — The limits of this genus remain to be defined; besides those
listed (the non-recognition of which will be justified below), several other
nominal genera may have to be united with it (see also Myers, 1928: 83).
Moreover, a number of species which evidently belong in this same group
have been randomly scattered by their describers over various genera with
Which they do not appear to have any close relationship, and even if correc¬
tions have been published, as in the case of Rhamdella montana Eigenmann,
1913, which, as Steindachner (1915) quite correctly pointed out, belongs in the
relationship of Chasmocranus = Heptapterus, this has been overlooked by
later compilers (Gosline, 1945; Fowler, 1951).

The genus ranges throughout the northern part of South America, to at
least as far south as the Rio de la Plata, but because of its uncertain limits

12

i;8

ZOOLOGISCHE VERHANDELINGEN 132 (I974)

neither its range nor the number of described species (presumably about
30) can be given with any accuracy.

Pariolius was re-described by Gosline (1940), who supplied it with the
following comparative diagnosis: “Of the five pimelodid genera besides
Pariolius without a free orbital rim and without dorsal or pectoral spines,
Pariolius may at once be separated from Heptapterus Bleeker, 1858, Acen-
tronichthys Eigenmann & Eigenmann, 1889, and Rhatndiopsis Haseman,
1911, by the longer anal fin (18 to 28 rays) in the last three genera. From
Chasmocranus Eigenmann, 1912, Pariolius differs in having a narrower
series of premaxillary teeth without backwardly projecting angles at the
sides and from Cetopsorhamdia Eigenmann, 1916, in lacking a forked caudal
and in having the ventral fins placed forward of the dorsal. Nannoglanis
is here considered a synonym of Pariolius.”

In a previous paper I have shown that neither the number of anal rays,
nor the shape of the premaxillary tooth patch are useful as generic characters
in this group, to which comes that in coloration and general appearance
Pariolius armillatus is a thoroughly typical Heptapterus. In my opinion, there¬
fore the species should be known as Heptapterus armillatus (Cope).

Nannoglanis was united with Pariolius by Gosline (1940), and therefore
also enters the synonymy of Heptapterus. Acentronichthys and Chasmocranus
were synonymized with Heptapterus by me (Mees, 1967).

Schultz (1944b) gave as characters of Imparales, compared with several
previously described nominal genera: “Among those pimelodid genera without
a free orbital rim, lacking spines in dorsal and pectoral fins, and with as few
as 12 anal rays, this new genus differs in having a forked caudal fin with
the upper lobe greatly elongate and the adipose fin confluent with caudal fin.
Rhamdiopsis Haseman, Acentronichthys Eigenmann & Eigenmann, and
Heptapterus Bleeker all have 18 to 28 anal rays, while the new genus has
hut 12. Chasmocranus Eigenmann has the premaxillary band of teeth with
backwardly projecting angles and the caudal fin not deeply incised. Pariolius
Cope has the caudal fin rounded and the pelvics inserted well in advance of
the dorsal origin, instead of a deeply incised caudal fin and pelvics inserted
under front of dorsal fin base as in Imparales .”

Here again the main characters given, premaxillary band without back¬
wardly projecting angles and shape of caudal fin, are amongst the most
variable found in the genus Heptapterus , and the two species currently placed
in Imparales should be known as Heptapterus mariai (Schultz) and Heptap¬
terus panamensis (Bussing).

The only character in which Medemichthys differs from sundry species
of Heptapterus is the very long upper caudal lobe; in view of the great varia-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

179

tion in the shape of the caudal fin in Heptapterus, I do not consider it suffi¬
cient to maintain a monotypic genus for this single character. The species
therefore becomes Heptapterus guayaberensis (Dahl).

In describing Phenacorhamdia, Dahl (1961) stated that: “this new genus
would seem to be somehow related to Chasmocranus Eigenmann but differs
in the projecting lower jaw, the pelvics inserted in front of the dorsal, and
the prolonged caudal lobe”. However, a slightly projecting lower jaw, as
described and figured by Dahl for P. macarenensis, occurs widely in Hep¬
tapterus, and pelvics implanted before the dorsal is a normal condition in that
genus. The one species described in Phenacorhamdia must therefore be known
as Heptapterus macarenensis (Dahl).

For a key to the five species known from Suriname, and their descriptions
and bibliographies, I refer to a previous publication (Mees, 1967); here only
such information will be given, as is additional to my earlier paper.

Heptapterus brevior (Eigenmann)

Chasmocranus brevior Eigenmann, 1912, Mem. Carnegie Mils., 5: 162, pi. 15 fig. 1 —
Waratuk, British Guiana.

Chasmocranus brevior ; Steindachner, 1915, Denkschr. Akad. Wiss. Wien, Math.-
Naturw. Kl., 93: 63 (no locality) ; Henn, 1928, Ann. Carnegie Mus., 19: 75 (Waratuk);
P. de Miranda Ribeiro, 1968, Bol. Mus. Nac. Rio de Janeiro, Zool., 262: 6 (no locality).

Heptapterus brevior ; Mees, 1967, Zool. Meded., 42: 217 (description and discussion).

Distribution. — No additional material has become available and the species
remains known from two localities only: Waratuk, Potaro River, British
Guiana (Eigenmann, 1912) and Nassau Gebergte, Suriname (Boeseman,
1953; Mees, 1967).

Discussion. — In my previous paper (Mees, 1967: 219) I mentioned
a specimen recorded as Chasmocranus brevior by Borodin (1927) from
Franca, Prov. Sao Paulo, Brazil. At the time I had not studied the specimen,
but on geographical grounds was inclined to reject the record as based on
misidentification. Through courtesy of Mr. J. N. Baskin, of the American
Museum of Natural History, I have now been able to examine the specimen,
and found my doubts of its identity confirmed. The specimen (AMNH
no. 9090), with a standard length of 29 mm, is superficially very similar
to H. brevior but differs in the characters listed on the next page.

I am unable to assign the AMNH specimen to any known species: it may
possibly belong to the recently described Heptapterus loped (P. de Miranda
Ribeiro, 1968b), but in that case the character mentioned under (1) must
be variable, as the published figure shows a fish in which the adipose fin
reaches dearly farther back than the anal fin. Also H. loped was described

i8o

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

as having eight dorsal rays. Heptapterus brachynema (Gomes & Schubart,
1958) differs in the characters (i) and (2). Anyway, the supposed occur¬
rence of H. brevior in south-eastern Brazil can now be definitely dismissed.

AMNH no. 9090

1) depressed A reaches nearer to
caudal base than does the
adipose fin.

2) V reaches nearly to A.

3) distance from tip of snout to
implantation of V: distance
from V to caudal base = 1: 1.1.

4) there are apparently more
vertebrae, but the vertebral
column is somewhat desinte-
grated, making an exact count
impossible; ca 42 (?).

H. brevior

depressed A reaches not so far
backwards as the adipose fin.

V reaches only about half way
(0.5 to 0.6) to A.
distance from tip of snout to
implantation of V: distance from V
to caudal base — 1: 1.5.
vertebrae 37.

Heptapterus longior (Eigenmann) (fig. 35)

Chasmocranus longior Eigenmann, 1912, Mem. Carnegie Mus., 5: 160, pi. 14 fig. 2 —
Amatuk, British Guiana.

Chasmocrantis longior ; Steindachner, 1915, Denkschr. Akad. Wiss. Wien, Math.-
Naturw. Kl., 93: 63 (Rupununi) ; Henn, 1928, Ann. Carnegie Mus., 19: 76 (Amatuk);
Grey, 1947, Fieldiana, Zool., 32: 130 (British Guiana: Amatuk, lower Potaro River);
Bohlke, 1953, Stanford Ichth. Bull, 5: 44 (Amatuk, Lower Potaro River, British
Guinea); P. de Miranda Ribeirt), 1968, Bol. Mus. Nac. Rio de Janeiro, Zool., 262: 6
(no locality); Lundberg & Baskin, 1969, Amer. Mus. Novit., 2398: 6, 28 (no locality).

Heptapterus longior ; Mees, 1967, Zool. Meded., 42: 220 (description and discussion).

Additional material. — Three specimens, 31 January 1925, Sao Gabriel Rapids, Rio
Negro, Brazil (C. Temetz, SU no. 58582), standard length 32, 32, 37 mm. One specimen,
11 December 1957, Awai Creek, a tributary of the upper Rupununi River near Isherton,
British Guiana (R. H. McConnell, BM no. 1971.7.29: 26), standard length 67^2 mm.
Four specimens, 15 December 1957, Arakwai Creek, just above inflow into the Rupununi
River, British Guiana (R. H. McConnell, BM no. 1971.7.29: 15-18), standard length
27, 36, 40, 64J/2 mm. Seven specimens, 1 December 1958, Rupununi, British Guiana (R.
H. McConnell, BM no. 1971.7.29: 19-25), standard length 43-76 mm. Two specimens,
30 April 1961, Rupununi, British Guiana (R. H. McConnell, BM no. 1971.7.29: 13-14),
standard length 65, 72 mm. Two specimens, 6 March 1964, Gran Kreek 1-2 km from
mouth (Boeseman, RMNH no. 26190), standard length 49, 92 mm. Three specimens,
6 March 1964, Gran Creek 4 km from mouth (Boeseman, RMNH no. 25562), standard
length 38, 40, 40 mm. One specimen, 13 August 1965, Mamadam, Suriname Rivier (Mees,
RMNH no. 26191), standard length 91 mm. One specimen, 18 October 1965, Compagnie
Kreek (Mees, RMNH no. 25765), standard length 54 mm. Six specimens, 6 February

MEES, AUCHENIPTERIDAE AND PIMELODIDAE l8l

1966, Sipaliwini (Mees, RMNH no. 26153), standard length 31-61 mm. One specimen,
12 October 1966, Sara Kreek, ca. 3 km S. from Dam (Nijssen, ZMA no. 105510),
standard length 43 mm. Five specimens, 14 October 1966, Sara Kreek, ca. 27 km S. from
Dam (Nijssen, ZMA no. 105295), standard length 38, 42, 54, 84, 95 mm, 17 specimens,
20 October 1966, Gran Kreek, 63 km S. from Affobakka (Nijssen, ZMA no. 105285),
standard length 31-74 mm. 32 specimens, 22 November 1966, Witte Kreek, 12 km S.
from Brownsweg (Nijssen, ZMA no. 105278), standard length 38-94 mm. 21 specimens,
8 December 1966, Marshall Kreek (Nijssen, ZMA no. 105371), standard length
29-74 mm. One specimen, 31 January 1967, tributary of Gran Rio, 4 km N. of Awaradam
(Nijssen, ZMA no. 105420), standard length 50 mm. One specimen, 1 February 1967,
tributary of Gran Rio, 3 km N. of Awaradam (Nijssen, ZMA no. 105546), standard
length 60 mm. 13 specimens, 27 February 1967, tributary of Kleine Saramacca, 11 km
from mouth of the latter (Nijssen, ZMA no. 105666), standard length 40-74 mm.
14 specimens, 28 February 1967, Kleine Saramacca, 14 km from mouth (Nijssen, ZMA
no. 105587), standard length 39-79 mm. Five specimens Suriname Rivier, ca. 754 km N.
from Botopassie (Nijssen, ZMA no. 105937), standard length 43, 66, 75, 79, 83 mm.
Two specimens, 20 March 1967, Parwapa Kreek near Botopassie (Nijssen, ZMA
no. 105794), standard length 58, 62 mm. Two specimens, 22 March 1967, Suriname
Rivier near Botopassie (Nijssen, ZMA no. 105788), standard length 68, 84 mm. 36 speci¬
mens, 5 April 1967, Stondansie, Nickerie Rivier (Nijssen, ZMA no. 105773), standard
length 2654-112 mm. Nine specimens, 5 April 1967, tributary of Nickerie Rivier, ca.
12 km S. of Stondansie (Nijssen, ZMA no. 105825), standard length 38-56 mm. 74 speci¬
mens, 6 April 1967, rapids in the Fallawatra Rivier, 5 km S. of Stondansie (Nijssen,
ZMA no. 105768), standard length 29-100 mm. 11 specimens, 21 April 1967, Maka
Kreek, tributary of the Lawa Rivier near its mouth (Nijssen, ZMA no. 105939), standard
length 51-91 mm. Seven specimens, 22 April 1967, Soeakisi Kreek, tributary of the
Tapanahoni near its mouth (Nijssen, ZMA no. 105940), standard length 50-117 mm.
One specimen, 24 April 1967, Kamaloea or Saloea Kreek, eastern tributary of the Maro-
wijne, French Guiana (Nijssen, ZMA no. 105942), standard length 57 mm. Four speci¬
mens, 10 May 1967, upper course of Coppename Rivier, at 3°5i / N, 56°45'W (Nijssen,
ZMA no. 105943), standard length 37, 47, 48, 49 mm. Three specimens, 17 May 1967,
upper course of Coppename Rivier, at 3°49'N, 56°57'W (Nijssen, ZMA no. 105938),
standard length 58, 69, 87 mm.

Distribution (fig. 35). — French Guiana: Kamaloea or Saloea Kreek,
eastern tributary of the Marowijne (first record for the country, see list of
material). Suriname: generally distributed, except the area inhabited by the
following species. British Guiana: known at present only from the Essequibo
River and its major tributaries, Potaro and Rupununi (cf. Eigenmann, 1912,
pi. LXXIII). Brazil: Sao Gabriel Rapids, Rio Negro.

Correction. — In my previous paper an error occurs in the list of material
of this species (Mees, 1967: 221). Through a printer's error, the 15th line
from the bottom is duplicated from the 6th line from the bottom. It should
read: “(no. 25531), standard length 46-90 mm. Four specimens, same data
(no.”

Heptapterus tapanahoniensis Mees (fig. 35)

Heptapterus tapanahoniensis Mees, 1967, Zool. Meded., 42: 223 — Tapanahoni, about
two kilometres downstream of its confluence with the Paloemeu.

182

ZOOLOGISCHE VERHANDELINGEN 132 (l 974 )

Fig. 35. The distribution of Heptapterus longior (dots) and Heptapterus tapanahoniensis (crosses). I, Sao Gabriel Rapids, Rio
Negro. 2, Amatuk. 3, Warraputa Falls. 4 Rupununi (several localities). 5, Avanavero Vallen, Kabalebo. 6. Sipaliwini. 7, Ston-
dansie, Nickerie Rivier. 8, upper course of Coppename. 9, Kleine Saramacca. 10, Suriname Rivier basin (many localities through¬
out the shaded area), it, Kamaloea or Saloea Kreek, 12, confluence of Tapanahoni and Lawa. 13, Tapanahoni near mouth of Paloe-

meu. 14 Apadron Soela, Paloemeu. 15, Ana Paike Kondre, Lawa.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

183

Additional material. — Seven specimens, 1 November 1966, Apadron Soela, Paloemeu
(J. P. Gosse, ZMA no. 109367), standard length 67-123 mm. 13 specimens, 18 November
1966, Lawa Rivier, downstream from Ana Paike Kondre (J. P. Gosse, ZMA no. 109371),
standard length 49-103 mm.

Distribution (fig. 35). — The species is known from the type series, col¬
lected in the Tapanahoni, from about two kilometres below to two kilometres
upstream of its confluence with the Paloemeu, and from Paloemeu and Lawa
as listed above.

Discussion. — When I stated that H. tapanahoniensis replaces H. longior
in the Tapanahoni Rivier (Mees, 1967: 229), I expected that H . tapanahonien¬
sis would be more widely distributed in the Marowijne basin, and although my
previous statement remains correct, it came as a surprise to me that Nijssen
found in the Marowijne and in the lower course of the Tapanahoni only H.
longior and not H. tapanahoniensis. The fact that the two occur in the same
river system, confirms my earlier view that H . tapanahoniensis is a separate
species and not a subspecies of H. longior . In this connection I should per¬
haps have drawn more attention to the difference in number of vertebrae
between the two species (39 in H. longior, 43 in H. tapanahoniensis ). The
situation as it appears to be now, with H. longior occurring in the lower
reaches of the Marowijne basin and H. tapanahoniensis higher up, is most
interesting and would merit a thorough ecological investigation. Note that
in those rivers where H. tapanahoniensis is not known to occur, H. longior
ranges upstream almost to the sources.

Heptapterus surinamensis Bleeker

Heptapterus surinamensis Bleeker, 1862, Versl. Meded. Kon. Akad. Wetensch., 14:
387 — Surinama.

Pimelodus mustelinus; (pt.) Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat.
Poiss. (4 0 ed.), 15: 123, 125 (Surinam).

Heptapterus surinamensis ; Steindachner, 1882, Denkschr. Akad. Wiss. Wien, Mathem.-
Naturw. Cl., 44 (1): 8 (discussion of type) ; Mees, 1967, Zool. Meded., 42: 225 (descrip¬
tion and discussion).

Additional material. — Three specimens, 21 March 1967, rapids in the Suriname
Rivier yy 2 km north of Botopassie (Nijssen, ZMA no. 105935), standard length 68, 75,
79 mm. 14 specimens, 22 March 1967, Suriname Rivier near Botopassie (Nijssen, ZMA
no. 105936), standard length 54-128 mm.

Distribution. — Confined to the Suriname River basin; exact localities are
restricted to the main river between Mamadam and Awaradam, and the Gran
Kreek, but it is likely that the type-specimen, merely labelled “Suriname”,
was obtained farther downstream.

184

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Heptapterus bleekeri Boeseman

Heptapterus bleekeri Boeseman, 1953, Zool. Meded., 32: 3 — Marowini Basin, Nassau
Mountains.

Heptapterus mustelinus; Vaillant, 1899, Bull. Mus. Hist. Nat. Paris, 5: 155 (le haut
de fleuve Carsevenne: la riviere Lunier ou la riviere Carnot = Lunier); Vaillant, 1900,
Nouv. Arch. Mus. Paris, (4) 2: 124, 127 (la riviere Lunier) ; Eigenmann, 1912, Mem.
Carnegie Mus., 5: 64 (French Guiana).

Heptapterus bleekeri ; Mees, 1967, Zool. Meded., 42: 227 (description and discussion).

Additional material. — One specimen, June-September 1897, la riviere Lunier, Conteste
franco-bresilien — Amapa, Brazil (F. Geay, MP no. 98-55), standard length 89 mm.

Distribution. — This species is now known from a creek in the Nassau
Gebergte, Marowijne basin, Suriname; the Tapoeripa Kreek near Broko-
pondo, Suriname Rivier basin, Suriname, and the Lunier, a northern tributary
of the Carcevenne or Calgoene, State of Amapa, Brazil.

Discussion. — Vaillant’s (1899, 1900) record of Heptapterus mustelinus
from Amapa, Brazil, far outside the accepted range of that species, appeared
to require verification, and examination of the specimen in question revealed
that it belongs to H. bleekeri. Vaillant may be excused for this misidentifica-
tion, for H. mustelinus and H. bleekeri are very similar. Two specimens of
H. mustelinus which I have examined, and which differ rather distinctly
from each other, can be distinguished from H. bleekeri by the following
characters: D 19 against D 20-22, P 9 (i.8) against P 8 (i.7), gill-rakers
10-14 against 6-8, predorsal length 2.85-3.0 times in standard length against
2.5-2.8, and origin of ventrals a little behind origin of dorsal (about opposite
2nd ray of dorsal), against slightly but distinctly in advance of dorsal origin.

Some characters of the specimen from Amapa slightly expand the range
of variation of H. bleekeri as given in my earlier publication; measurements
and counts of this specimen are: D 7 (i.6), A 20, P 8 (i.7), V 6 (i.5), C 13
(branched rays only), gill-rakers 8, branchiostegals 8, predorsal length 2.5
times in standard length, distance of origin of A to tip of snout 1.55 times
distance of origin of A to caudal base, adipose base 3.0 times in standard
length. In my previous publication I mistakenly gave the number of divided
caudal rays as 20 to 22; this should read developed caudal rays: the number
of divided rays is only 13 or 14.

Heptapterus somnians species nova (fig. 36)

Material. — One specimen, 2 May 1968, Sangadina, a tributary of the Rio das Mortes
2 km from Xaventina, Mato Grosso, Brazil (R. H. McConnell, BM no. 1971.7.29: 4),
standard length 55 mm, holotype of the species.

Characters. — A typical member of its genus, characterized by a short anal
fin, deeply forked caudal fin, a remarkably broad band of teeth in the lower
jaw, and uniform coloration.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

185

i86

ZOOLOGISCHE VERHANDELINGEN 1 32 (1974)

Description. — D 7, A 10, P 8, V 6, C i.15.1, besides shorter simple rays
and rudiments; number of vertebrae 39; gill-rakers on the outer branchial
arch 6, rather close together on the upper half of the hypobranchial, well-
developed, but decreasing in size downwards, and the last one small; bran-
chiostegals 7, well-developed; mouth terminal; lips fleshy, the lower lip with
8 pores; teeth villiform, depressible, in the upper jaw in a single broad band,
without backward projections, in the lower jaw in a band which is quite as
broad as that of the upper jaw (fig. 36a); eyes small, difficult to measure,
but their greatest diameter about 4 times in snout, 2.5 in interorbital, and 11
in head; the two pairs of nostrils forming a square; the upper surface of the
head is covered with numerous minute papillae; the maxillary barbels reach
to the end of the opercle; both pairs of mandibular barbels are of almost
identical length, and attain about two-thirds of the length of the maxillary
barbels.

Head 4.8 times in standard length, greatest width of body 7.5 in standard
length, depth of body 10.2 in standard length, predorsal length 2.7 in standard
length, distance from tip of snout to base of ventrals 2.8 in standard length.

Dorsal fin with one simple and five divided rays, and one rudiment, its
base 9.4 in standard length, longest dorsal ray about two-thirds length of head,
or equal to head without snout 1 ). Anal fin with ten rays, the first three of
which are simple, the others divided; distance of origin of A to tip of snout
2.0 times its distance to base of caudal fin. Pectorals more or less rounded,
with one simple and seven divided rays, their second and third rays longest,
equal to head without snout, also equal in length to longest dorsal rays.
Ventrals rounded, with one simple and five divided rays, their origin slightly
in advance of origin of D; ventrals equal in length to the pectorals. Adipose
fin originating opposite the fifth ray of A, its height about half the depth of
the body below it, not nearly reaching to, and widely separated from, the
caudal fin; length of its base 5.6 in standard length. Caudal fin deeply
forked, its lobes narrow, but rounded at the tips, the lower lobe slightly the
larger and longer; longest rays twice the length of the longest pectoral rays,
central rays two-fifths of the length of the longest rays; there are 15 divided
rays, flanked above and below by a fully developed simple ray, and by many
smaller and rudimentary simple rays.

1) The usual number of dorsal rays in the genus is seven (i.6). In the present speci¬
men I counted originally i.5 only, but the X-ray photograph showed clearly seven rays,
of which the fifth, however, was shorter than the others and simple. By removing the
thick skin covering the base of the fin, beyond which it did not protrude, I found the
missing ray. It is evident that this situation is aberrant and that the normal fin-formula
is i.6 (and not L3.L2 as in this specimen).

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

187

Colour. Even earth brown, the upper surface of the head darker, blackish
brown, the fins lighter, with the pigment moreover restricted to the rays, the
membranes being hyaline, except in the adipose fin which has especially its
basal part well-pigmented.

Genus Pseudopimelodus Bleeker

Pseudopimelodus Bleeker, 1858, Visschen Ind. Arch., 1: 196, 204 — type by sub¬
sequent selection (Bleeker, 1862, Atlas Ichth., 2: 11), Pimelodus raninus Valenciennes.

Zungaro Bleeker, 1858, Visschen Ind. Arch., 1: 196, 204 — type by original designation
and monotypy, Pimelodus zungaro Humboldt.

Batrachoglanis Gill, 1858, Ann. Lyc. Nat. Hist. New York, 6: 389 — type by original
designation, Pimelodus raninus Valenciennes.

Cephalosilurus Haseman, 1911, Ann. Carnegie Mus., 7: 317 — type by monotypy,
Cephalosilurus fozvleri Haseman.

Diagnosis. — A genus of probably rather sluggish bottom fishes with large
and heavy depressed heads, entirely covered with thick skin; the head is
broadest between the opercles, and from there the body tapers towards the
tail; eyes small, covered with skin, situated on the dorsal surface of the
head, hence looking upwards, not sideways; teeth in each jaw in a band, in
the upper jaw with lateral projections backwards, no teeth on vomer and
palatines; postoccipital process well-developed, meeting or almost meeting
the predorsal plate; barbels short or of moderate length, the maxillary barbels
vary from not quite reaching the base of the pectorals, to reaching just
beyond the base of the dorsal spine; pectoral pore present, below the middle
of the postcleithral process, in one species (P. zungaro ), absent in the other
members of the genus; dorsal spine short, either with teeth or smooth;
pectoral spines longer (but less than half the length of the head in the species
here dealt with), with well-developed teeth or hooks on both margins (rarely
along the posterior margin only); anal fin short, with seven to nine divided
rays; adipose fin short but well-developed, rounded; caudal fin variable in
shape: forked with pointed or with rounded lobes, truncate or lanceolate.
Lateral line complete, but usually inconspicuous, arched in its anterior part
or almost straight. Colour dark brown to blackish, with a pattern of pale, light
brown or white bands and streaks, most distinctive in small specimens; fins
often largely black, usually with white bands.

Distribution. — The genus has a wide distribution in South America,
from Colombia and Venezuela through the Amazon basin to Uruguay; one
species occurs in western Colombia and Ecuador, on the Pacific side of
the Andes.

Discussion. — The genus Zungaro was established for Pimelodus zungaro
Humboldt. This species was based on a description and plate only, the type

l88 ZOOLOGISCHE VERHANDELINGEN 132 (1974)

was not preserved, and as mentioned below and in the discussion of P.
bufonius, the name Zungaro zungaro has often been misapplied.

For an identification of Z. zungaro we have to rely entirely on Humboldt’s
description and plate (pi. 5). Consulting these, one finds that neither its
colour and colour pattern (blackish olive with numerous small black ocelles),
nor its fin-ray formula (D 7, P 13, V 10, A 10, branchiostegals 4), comes
anywhere near a known species of Pseudopimelodus. Its size (3 pieds 4 pou-
ces) is also larger than any member of Pseudopimelodus except P. man-
gurus which was supposed to attain a length of over six feet (Valenciennes,
1840: 117), although Steindachner (1880) recorded under the name P. bufo¬
nius a specimen of 55 cm. Valenciennes (1840: 119) wrote: “M. de Hum¬
boldt ... dit toutes les nageoires sans piquans; mais il n’a probablement pas
disseque les membranes epaisses qui (ajoute-t-il) recouvrent ces parties au
point que l’on a peine a en sentir les os”. Even though Valenciennes was
probably right in his criticism, the number of discrepancies remains such
that Z. zungaro cannot be easily identified with any species now known.

Eigenmann (1922: 32) stated that the name “Zungaro” is applied to
various large silurids in the Peruvian Amazon, so that this does not help
either.

It is apparent that much confusion exists about the validity and characters
of Zungaro. In this connection I want especially to mention the discrepancies
in the description of the genus and its type species by various authors (tail
rounded or forked, teeth along the anterior edge of the pectoral spine present
or absent, or present in small specimens, but disappearing with age). From
Gosline’s review Zungaro has been omitted. Evidently several species have
been confused under this name and from descriptions alone these cannot
always be identified. The validity of Zungaro depends clearly on the correct
identification of its tvpe-species. For a considerable time I have held the
opinion that Pimelodus zungaro Humboldt was unidentifiable, and should for
that reason be placed as a species inquirendae. Gradually, however, I have
come round to the opinion that the species can be identified, for, however poor
Humboldt’s description is by modern standards, it does agree much better
with the species here listed under the specific name zungaro , than with any
other fish known from the Amazon basin. The absence of strong armature,
combined with the huge size, appears to rule out any other known species and
it is also unlikely (although possible) that Humboldt’s specimen would have
belonged to a species never since recorded. With the identity of P. zungaro
settled, the question of the validity of the genus Zungaro as different from
Pseudopimelodus remains. Recognition of genera has always a subjective
element, but it is my opinion that when defining genera, similarities rather

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

189

than differences should be stressed. In the following pages evidence will be
presented that if Zungaro is recognized, this genus will be monospecific. The
genus Pseudopimelodus as defined by me, including Zungaro and Cephalosilu-
rus, counts only five spedes, and within this group, the resemblances between
the species are much greater than their differences. P. zungaro fits well in
the genus Pseudopimelodus, although, admittedly, certain characters stamp
it as the most distinctive of its genus.

Fig. 37. Dentition of Pseudopimelodus. (a) P. zungaro bufonius, (b) P. r. raninus ,
(c) P. albomarginatus , (d) P. nigricauda.

When Haseman (1911) described Cephalosilurus, he compared it with
Lophiosilurus, but made no mention of Pseudopimelodus. There is nothing in
the description of Cephalosilurus fowleri that would preclude its being placed
as a normal representative of the genus Pseudopimelodus. The only differ¬
ence given by Gosline (1941) to distinguish between Cephalosilurus and
Pseudopimelodus is that the former has: “Lower jaw projecting beyond
upper; fore-part of head broad and greatly flattened”, and the latter: “Lower
jaw not projecting beyond upper; fore-part of head not excessively flattened”.
However, the head of P. raninus, the type species of Pseudopimelodus, is also
flattened. On the basis of Gosline’s diagnosis, the species described on a later
page as Pseudopimelodus nigricauda would belong to the genus Cephalosilu¬
rus. Moreover, P. bufonius, placed in Cephalosilurus by Gosline (1941, 1945),
has an only slightly projecting lower jaw, and its head is no more flattened
than that of P. raninus (incidentally, P. bufonius is conspecific with P. zun¬
garo, type of the genus Zungaro which antedates Cephalosilurus by over fifty
years). It is evident that both characters on which Gosline relied for the
distinction of genera, vary from species to species, and the same can be said
of the dentition which has also been used for generic discrimination (fig. 37).

190

ZOOLOGISCHE VERHANDELINGEN 1 32 (1974)

When P. parahybae is removed to the genus Microglanis, as first proposed
by Gosline (1941), there remain fourteen nominal speries and one subspecies.
Various authors have, from time to time, expressed doubt about the validity
of several of these species, and the situation has become somewhat confused,
for which reason it appeared worth while to attempt a revision based on as
much material as it was possible to borrow and beg together. Nevertheless,
the results are in many respects unsatisfactory. I have had to leave several
problems unsolved, and in other instances the solutions offered will probably
have to be modified when more material becomes available.

Thus I have arrived at the following tentative arrangement.

1. Pseudopimelodus zungaro, with the subspecies:

P . zungaro bufonius (synonym : P. charus ).

P. zungaro mangurus (synonym: P . roosevelti).

P. zungaro zungaro (synonyms: P . pulcher, P . variolosus) .

P. zungaro subsp.

2. Pseudo pimelodus raninus, with the subspecies:

P . raninus raninus (synonym: P. acanthochirus) *).

P. raninus villosus.

P. raninus acanthochiroides (synonym : P. villosus butcheri ).

P. raninus transmontanus.

3. Pseudo pimelodus fowleri.

4. Pseudo pimelodus albomargniatus

5. Pseudo pimelodus nigricauda.

Hitherto four nominal species of the genus Pseudo pimelodus had been
recorded from the Guianas: P. raninus from the Mana, Cayenne (Valen¬
ciennes, 1840) and Litani, Suriname (Hoedeman, 1961), P . bufonius from
Cayenne (but see discussion of its type-locality), P. albomarginatus and
P. villosus from British Guiana (Eigenmann, 1912: 154; Lowe, 1964: 116,
140). Hoedeman (1961) stated that his specimens of P. raninus were from
French Guiana, “Litany river, village Aloike”, but the village Alowike (as
it is spelled on our maps) is definitely in Suriname, not French Guiana.
Eigenmann (1912: 154) mentioned that specimens from the Essequibo
recorded by Gunther (1864) under the name Pimelodus raninus do not belong
to that species, but he failed to give an opinion on their identity. I have
examined these specimens and found them referable to P. villosus; they had
already been labelled with that name, presumably by Regan.

1) This name has been changed to Pseudo pimelodus acanthochira , or even P. acantho -
cheira by subsequent authors, beginning with Eigenmann & Eigenmann (1890: 114) them¬
selves. No reason for this change has been given, and none is obvious.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE 191

Our recent collections from Suriname contain four forms (three species)
of the genus Pseudopimelodus, and obviously these had to be compared with
the species already known from the Guianas. One of the species agrees well
with the description and especially the plate given by Valenciennes of P.
raninus, with which it can be identified; the second appears to be referable to
P. bufonius ; the third is P. villosus (here regarded as a subspecies of P.
raninus), and the fourth is new.

Fig. 38. Gill-rakers on outer branchial arch of Pseudopimelodus. (a) P. zungaro bufonius ,
(b) P. r. raninus , (c) P. albomarginatus , (d) P. nigricauda.

Key to the species of Pseudopimelodus *)

ia — Pectoral fins usually with seven rays (sometimes six in small speci¬
mens), pectoral spine entirely encased in thick skin, anal fin with six
or seven divided rays, maxillary barbels reaching to base of pectoral
fin or a little beyond, dorsal fin when pressed backwards reaching
no more than about half way to adipose fin, occipital process and dorsal
plate in broad contact, caudal fin falcate, white or dusky with a black
cross band and a black base, or mostly dark, gill-rakers little developed,
1 + 1 + 1 to S, anterior nostril some distance behind the upper lip

. P. zungaro

b — Pectoral fins with six rays, pectoral spine covered with thin skin only,
which leaves the spine and its teeth clearly visible, dorsal fin when
pressed backwards usually reaching well beyond half way to the
adipose fin, occipital process and predorsal plate separated or in con¬
tact, caudal fin scalloped, rounded, or lanceolate, never falcate, gill-
rakers on outer branchial arch at least io, long or short (or rudimen¬
tary), anterior nostril on upper lip.2

1) It deserves mention that the gill-rakers are implanted not in a single series but
often in a double series; this is particularly conspicuous in P. albomarginatus, in which
the inner series is as well-developed as the outer series. In P. nigricauda there are even
a few papillae carrying three rakers. In the figures as well as in the counts, only the
outer series has been considered.

192

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

2a — Only three or four gill-rakers on the outer branchial arch well-
developed, all others, some ten in a row on the hypobranchial, ex¬
tremely short, rudimentary (fig. 38d), caudal fin rounded or lan¬
ceolate .3

b — At least ten gill-rakers on the outer branchial arch well-developed, and
only the last one or two rudimentary (fig. 38b and c), caudal fin

scalloped, rounded or more or less truncate.4

3a — Fins not very long: the dorsal fin when depressed covers only about
three-fifths of its distance from the adipose base, pectoral spine 2.9
times in predorsal length, ventrals not reaching to anal origin, caudal
fin rounded, its greatest length about 4.5 times in standard length,

caudal fin spotted, not black. P. fowleri

b — Fins long: the dorsal fin when pressed backwards reaches, or almost
reaches the adipose fin, pectoral spine 2.5-2.6 times in predorsal length,
ventrals reaching or almost reaching to anal origin, caudal fin
rounded to lanceolate, its greatest length less than four times in stan¬
dard length, caudal fin black, with a very narrow white margin

. P. nigricauda

4a —- Head large, 2.5-2.7 times in standard length, predorsal length 2.2-2.35
times in standard length, the maxillary barbels do not reach, or only
just reach to the pectoral base, premaxillary band of teeth with large
backward projections laterally, caudal fin rounded-lanceolate

. P. albomarginatus

b — Head moderate, 2.9-3.0 times in standard length, predorsal length
2.4-2.5 times in standard length, the maxillary barbels usually reach
beyond the pectoral base, premaxillary band of teeth with moderate
backward projections laterally, caudal fin scalloped, in large specimens
often more or less truncate. P. raninus

Pseudopimelodus zungaro (Humboldt)

The main characters of this species appear in the key, and descriptions
are given under the subspecies P. z. bufonius and P. z. mangurus.

The presence of teeth on the anterior margins of the pectoral spines appears
to be a matter of size; teeth are conspicuous in all small specimens, but
become obsolete in individuals with a standard length of 200 mm and over.
Schultz's (1944a: 197) generalisation that: “Zungaro ... differs from the
related genera, Pseudopimelodus and Microglanis, in having ... the anterior
margin of the pectoral spine smooth” is only partially true in his own mate¬
rial as from sample USNM 121283 only the two larger specimens lack the
teeth; in the two smaller ones they are clearly present. In Schultz's (1944a:

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

193

186) key the first difference between Pseudopimelodus and Zungaro is
“anterior edge of pectoral spine serrated”, against “anterior edge of pectoral
spine smooth”, so that one is bound to come to grief.

P. zungaro can attain a very large size. Although the largest specimen I
examined had a standard length of only 257 mm, in literature measurements
of up to six feet have been recorded. Possibly these were exaggerated, and
reliable measurements and weights of large individuals remain to be taken.

The species is widely distributed in northern and eastern South America,
from the Magdalena basin in Colombia, through Venezuela and the Guianas,
the whole Amazon basin, the Rio Sao Francisco and the Parana- La Plata
basin.

Within its huge range, P. zungaro shows some slight geographical variation
in colour-pattern, on the basis of which I have recognized four subspecies. It
should be noted that these subspecies are not very satisfactory, and the main
reason why I accept them is that they have been described and named long
ago. If this had not been the case, I would probably have described the varia¬
tion, without applying trinominals. For the same reason I have also left the
subspecies from the Rio Magdalena without a name, although it appears to
be the most isolated and a reasonably distinctive one.

Pseudopimelodus zungaro bufonius (Valenciennes)

(figs. 37a, 38a, frontispiece, pis. 6, 7)

Pimelodus bufonius Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.), 35: 115 — Cayenne (but probably Suriname).

Pimelodus charus Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.), IS: 118 — Rio Sahara.

Pseudopimelodus bufonius ; Bleeker, 1858, Visschen Ind. Arch., x: 207 (Guyana);
Liitken, 1875, Vidensk Selsk. Skr., (5) 12: 180 (Cayenne); (pt.) Steindachner, 1880,
Denkschr. Akad. Wiss. Wien, Mathem.-Naturw. Cl., 42 (1) : 59 (Rio das Velhas);
Eigenmann, 1912, Mem. Carnegie Mus., 5: 151, 152 (no locality).

Pimelodus bufonius ; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 133 (Rio Cipo, a tributary
of the Rio das Velhas, Minas Geraes).

Pseudopimelodus charus ; Bleeker, 1858, Visschen Ind. Arch., 1: 207 (Amer. merid.) ;
Liitken, 1875, K. Danske Vidensk. Selsk. Skr., (5) 12: 118 (Rio das Velhas).

Pseudopimelodus zungaro ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) 1: 122 (Goyaz) ; Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1:112
(Goyaz) ; A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 254, fig.
105, pl. 43 fig. 1 (rio S. Francisco).

Pseudopimelodus zunigaro ; (pt.) Eigenmann & Eigenmann, 1891, Proc. U.S. Nat.
Mus., 14: 28 (Rio Plata to Rio Magdalena).

CepJtalosilurus bufonius ; Gosline, 1941, Stanford Ichth. Bull., 2: 85 (Cayenne) ; Gosline,
1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 27 (Caiena).

Zungaro zungaro ; Schultz, 1944, Proc. U.S. Nat. Mus., 94: 197 (Rio San Juan at the
bridge south of Mene Grande, tributary Rio Motatan and Rio Negro below mouth Rio
Yasa, Maracaibo Basin of Venezuela) ; (pt.) Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro,

13

194

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Zool., 33: 33 (Venezuela); Dahl, 1961, Noved. Colomb., 1: 490 (Guaybero River, Orinoco
System, Colombia); Mago-Leccia, 1967, BoL Soc. Venez. Cienc. Nat., 27: 255 (Vene¬
zuela).

Zungaro charus ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 33 (Brasil).

Zungaro bufonius ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33 : 33 (Caiena).

Material. Eight specimens from Suriname as follows. One specimen, 3/4 February
1964, Suriname Rivier at Brokopondo (Boeseman, RMNH no. 26604), standard length
73 mm. One specimen, 17 July 1965, Awaradam, Gran Rio (Mees, RMNH no. 26605),
standard length 102 mm. Three specimens, 23 January 1966, Sipaliwini (Mees, RMNH
no. 26606), standard length 120, 133, 205 mm. One specimen, 6 February 1966, Sipaliwini
(Mees, RMNH no. 26607), standard length 104 mm. One specimen, 5 April 1967, Ston-
dansie, Nickerie Rivier (Nijssen, ZMA no. 105945), standard length 243 mm. One speci¬
men, same data (Nijssen, ZMA no. 105946), standard length 76 mm.

Extra-limital material. — One specimen, 1850-1856, Rio das Velhas, Brazil (J. Rein¬
hardt, UZMK no. 330), standard length 202 mm. Two specimens, same data (UZMK
nos. 333, 334), standard length 119, 131 mm. Four specimens, no date, Goyaz (Honario,
MCZ no. 8196), standard length 116, 132, 187, 202 mm. Four specimens, 20 March 1942,
Rio San Juan at bridge, Maracaibo basin, Venezuela (L. P. Schultz, USNM no. 121283),
standard length 152, 173, 197, 257 mm. One specimen, 1942, Pirapora, in rapids of Rio
Sao Francisco, Minas Gerais, Brazil (G. S. Myers, Carvalho, and P. de Miranda Ribeiro,
SU no. 64100), standard length 135 mm. One specimen, 25 April 1961, Rupununi near
Arakwa mouth, British Guiana (R. H. McConnell log 616), standard length 119 mm. One
specimen, 28 April 1961, Sandcreek pools, Rupununi District (R. H. McConnell log 630-
633), standard length 87 mm. One specimen, 30 April 1961, Sandcreek, Rupununi District
(R. H. McConnell log 649), standard length 90 mm. Two specimens, May 1961, Wichabai,
Rupununi River (S. Singh, Mrs. McConnell’s collection), standard length 66, 76 mm.

Diagnostic characters. — Head not so broad and depressed as in the other
species, rounded at the gape, its colour dark, mottled, with a broad pale nuchal
band, pectoral fins with seven rays (different from all other species of the
genus), maxillary barbels reaching to base of P or a little distance beyond,
the dorsal fin when depressed reaches only about half way to the adipose fin;
the caudal fin is falcate, pale with a blackish base and one blackish cross-bar;
occipital process and dorsal plate are in contact.

Description. — D I. 6, A iii to iv.6 to 7, P I.7 (sometimes 1 . 6 ), V i.5, C with
14 or 15 divided rays (one specimen has only n divided rays; this is evi¬
dently aberrant); gill-rakers very few 1 + 1 + 1 to 3; branchiostegals 8 or 9;
teeth in both jaws in a broad band, the band of premaxillary teeth with acute
projections laterally, which, however, do not reach far backwards (fig. 37a);
eyes small, their greatest diameter 2.5-3.5 times in snout, 2.5-5.2 times in bony
interorbital, and 9-12 times in head; the maxillary barbels reach to the base
of the pectoral fins, or beyond, almost to halfway along the pectoral fins; the
outer pair of mental barbels is about half their length, and the inner pair is
about half the length of the outer pair.

Head 3.3-3.7 times in standard length, greatest width of body 3.5-3.8 times
in standard length; greatest depth of body 4.6-57 times in standard length;

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

195

predorsal length 2.8-3.0 times in standard length (one specimen of sample
USNM no. 121283, has head 3.9, width 4.0, predorsal length 3.15 times in
standard length).

Dorsal fin with one spine and six divided rays, the spine moderately strong
and not very long, though always longer than snout with eye, encased in
leathery skin, which conceals the strong antrorse serrations it has on both
sides; soft rays comparatively short, when pressed down they reach only
about half way to the origin of the adipose fin. Anal fin with three or four
simple and six or seven divided rays, rounded, its longest rays about half
the length of the head. Pectorals with one spine and seven rays; the spine is
strong, in small specimens it is on each side and over its whole length provided
with about a dozen strong antrorse hooks, which are not so long as the width
of the spine, so that the spine as a whole is more slender than that of P.
nigricauda ; in specimens Of over 200 mm standard length, the teeth along
the anterior edge may be obsolete; the spine is encased in very thick leathery
skin, which is continued beyond its tip; it is most peculiar that also in the
leathery part, knobs appear on the posterior surface corresponding with the
hooks of the bony part; the spine with its protruding soft part, is longer
than any of the rays; the length of the spine is slightly more than snout with
eye; the length of spine with soft part equals the head without snout or head
without snout and eye; the rays become gradually shorter, and the last one
equals snout with eye. Ventrals rounded, with six rays, of which the first
one is simple, the others divided, reaching from two-thirds to three-fourths of
the distance to the anal fin, the longest rays equal in length to the longest
rays of the anal fin. Caudal fin well-developed, forked to falcate, always with
acute tips, with n-15 divided rays besides several developed simple rays
and some rudiments, the longest rays very slightly shorter than the head.
Adipose fin of moderate length, its origin opposite of or slightly in advance
of the origin of the anal fin, and its base reaching backwards as far as the
base of the anal fin or very little beyond, its height about half the depth of
the caudal peduncle below it.

Colour. The colours are whitish to pale brownish and black or brownish
black. The head is fairly dark, mottled; there is a broad dark band across
the nape followed by an even broader pale band, which comes to just in
advance of the origin of the dorsal fin; the body has three broad, irregular
black bands; one in which the dorsal fin is placed; one in which the adipose
fin and the anterior part of the anal fin are placed, this band dorsally nar¬
rowly connected with the anterior one; and one on the caudal base. The dorsal
fin is blade, with the skin containing the spine paler, and the tips of the rays
whitish; there is an ill-defined white area half way up the posterior five rays

196

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

and their connecting membranes. The anal fin is whitish with a broad black
band across its base, beginning just beyond its origin, and a second slightly
narrower band across its distal half, leaving a pale tip, central band, and
origin. The pectorals are pale, with a dark cross-band, and basally more or
less mottled. The ventrals are similar in appearance to the pectorals. The tail
is pale with, as mentioned above, a conspicuous black base and a single fairly
narrow dark cross-band. The adipose fin is for its greater part contained
in the dark second body-band, but at the origin of its base is a small round
pale spot, and its whole posterior third is pale.

Habitat. — All our specimens from Suriname were taken in rivers, where
they appear to be not very common.

Distribution. — North-eastern South America from the Maracaibo basin,
Venezuela, through British Guiana (Rupununi) and Suriname to eastern
Brazil: Goyaz and the Rio Sao Francisco basin (Rio das Velhas and its
tributary the Rio Cipo). No certain record from French Guiana (the type-
locality Cayenne is doubtful, see Discussion).

In Suriname known from the Suriname Rivier, the Sipaliwini (Corantijn
basin) and the Nickerie Rivier.

Fowler (1951: 561) has mistakenly listed Steindachner (1876b: 632 ; 1880:
60) in the bibliography of Pseudopimelodus charus, thus adding the Rio
Paraiba and Santa Cruz to the distribution of this nominal species. It appears
to have escaped his attention that the fishes which Steindachner (1876b:
632) recorded under the name P. charus are the same ones which that author
described a few years later as P. parahybae. Fowler’s reference to Stein¬
dachner (1880: 60) makes even less sense, as on the place indicated, Stein¬
dachner clearly states that he regards P. charus as a synonym of P . bufonius,
whereas Fowler’s remark: “localidades presedentes” indicates continued con¬
fusion with P. parahybae. Fowler’s confusion is difficult to understand as
in the bibliography of Microglanis cottoides = P. parahybae, he gives the
same references to Steindachner under the correct species.

Discussion. — The use of the name P. bufonius requires some explanation.
The species was described from a single specimen with a length of “sept
pouces” in the Leiden museum. At present the type can no longer be found
in Leiden, and as it is not in Paris either (Bauchot, in litt. 27.II.1970; also
it was not listed by Bertin & Esteve, 1950), it may be considered lost. The
fact that neither Bleeker (1864) nor van der Stigchel (1946) made mention
of it, suggests that the specimen has disappeared a considerable time ago.

Since its description, P. bufonius has not, to my knowledge, been recorded
from anywhere near its type-locality. Many authors have speculated on the
affinities of P. bufonius, and without much justification (as they had no

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

197

material), shifted it from genus to genus. Ultimately the situation could
become so confused that Gosline (1945) listed it twice; on page 27 as
Cephalosilurus bufonius, and on page 33 as Zungaro bufonius.

Gomes (1946: 6) considered P. bufonius to be of doubtful validity, an
opinion for which he referred to Eigenmann (1912: 151); on the place
indicated, however, Eigenmann did not express doubt about the validity of
P. bufonius, but on the contrary, re-instated it, as quoted below. Eigenmann
& Eigenmann (1890: 112) placed P. bufonius as well as P. mangurus in the
synonymy of P. zungaro, and described under the last-mentioned name speci¬
mens from Goyaz. Later, Eigenmann (1912: 151) retracted his earlier state¬
ments : “In 1890 I identified Pimelodus bufonius with the Pimelodus zungaro
of Humboldt. Humboldt’s species is known only from his figure and descrip¬
tion, which are said to have been taken from a fish three feet four inches
long and reported to attain seven feet. It is very doubtful whether the zungaro
of Eigenmann and Eigenmann from Goyaz is identical with Humboldt’s
species. It is also very doubtful whether Pimelodus mangurus Valenciennes
from the La Plata is synonymous with zungaro. This elimination of synonyms
leaves the genus Pseudo pimelodus, as here understood, consisting of raninus
Cuvier and Valenciennes (the type), bufonius Cuvier and Valenciennes,
acanthocheira Eigenmann and Eigenmann, cottoides Boulenger, and the two
species here described”.

For a judgement of identity, appearance and affinities of P. bufonius , the
only source of information is therefore Valenciennes (1840). I have care¬
fully compared his description of the type with one of my specimens from
Suriname of comparable size (RMNH no. 26606, standard length 133 mm),
and found a close agreement, except that in the type : “Les epines de la dorsale
et des pectorales sont assez fortes, tres-finement dentelees au bord posterieur,
et ont tout leur moitie superieure nolle et articulee”. The spines of my speci¬
mens are entirely encased in leathery skin, providing soft flexible tips as
described by Valenciennes, but the spines have distinctive hooks along their
anterior borders. However, whereas in small specimens these teeth are
dearly visible, in the individual of 133 mm, the anterior rows of teeth are
entirely concealed under a thick layer of skin, and are invisible without
removing this skin. In view of the complete agreement otherwise existing
between Valenciennes’s description and my specimens, I believe that Valen¬
ciennes has overlooked the teeth on the anterior edges of the spines, for which
there is certainly some justification as explained above (in specimens of the
size mentioned by Valenciennes the teeth should be present). An additional
argument for accepting this identification is geographical: although Valen¬
ciennes described P . bufonius from Cayenne, there is every reason to assume

198

ZOOLOGISCHE VERHANDELINGEN 1 32 (1974)

that the type-specimen came actually from Suriname, and had been collected
by Dieperink, who lived in Paramaribo (cf. p. 43). The type-locality would
probably be the Suriname Rivier basin. The Suriname Rivier is now
ichthyologically reasonably wel known, and as one of the three species of
Pseudopimelodus occurring in it corresponds closely with the description of
P. bufonius, I see no reason to doubt its identity with that species.

P. charus was based on a drawing of a fish from the Rio Sahara. Sub¬
sequently Liitken (1875) recorded specimens from the Rio das Velhas under
that name. I am able to reproduce the drawing of P. charus here, thanks to
Madame Bauchot, who informed me as follows: “J’ai trouve ce dessin dans
les manuscrits de Valenciennes. Cest une aquarelle assez laide, dans les
tons gris-brun, aussi ai-je demande que la photographie soit en noir et
blanc”. The Rio Sahara equals the Rio das Velhas. As there is a good agree¬
ment between the drawing and Liitken’s three specimens, it is practically
certain that the latter were correctly identified as P. charus, of which they are
virtual topotypes. Comparison of these specimens with P. zungaro bufonius
from Suriname shows that they are scarcely separable by having an even
slightly shorter dorsal fin, which, when adpressed, reaches less than half way
to the adipose fin, whereas the white band across the middle of the dorsal
fin appears to be wider and more distinct. Although it is just possible that
in future, if more and fresh material is compared, the name charus can be
revived as a subspecies, I see for the present no reason to treat it as anything
but a synonym of P. zungaro bufonius.

Pseudopimelodus zungaro mangurus (Valenciennes)

Pimelodus mangurus Valenciennes, 1834-1837, in d’Orbigny, Voy. Amer. Merid., 9
(Atlas) : Poiss. pi. 1, fig. 4-6 — no locality, plate only, without description (description
and locality provided by Valenciennes, 1840, see below).

Pseudopimelodus roosevelti Borodin, 1927, Amer. Mus. Novit., 266: 1-3, fig. 1 —
Parassununga, Province of Sao Paulo, Brazil.

Pimelodus mangurus ; Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.), 15: 116 (depuis les Missions jusqu’au-dessus du confluent du Parana et de la
Plata); Gunther, 1864, Cat. Fish. Brit. Mus., 5: 134 (La Plata River).

Pimelodus bufonius ; Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw. CL, 26
(1857): 421 (Cujaba-Fluss); Goeldi, 1898, Bol. Mus. Paraense, 2: 456 (Cuyaba).

Pseudopimelodus mangurus; Bleeker, 1858, Visschen Ind. Arch., 1: 207 (Amer. merid.).

Pseudopimelodus bufonius; (pt.) Steindachner, 1880, Denkschr. Akad. Wiss. Wien,
Mathem.-Naturw. CL, 42 (1) : 59 (Cuyaba; La Plata).

Pseudopimelodus zunigaro; (pt.) Eigenmann & Eigenmann, 1891, Proc. U.S. Nat.
Mus., 14: 28 (Rio Plata to Rio Magdalena).

Pseudopimelodus zungaro; (?) Eigenmann & Norris, 1900, Rev. Mus. Paul., 4: 350
(S. Paulo, without exact locality); Fischer, 1917, Ann. Carnegie Mus., 11: 411 fig. 3
(Piracicaba); (pt.) A. de Miranda Ribeiro, 1918, Rev. Mus. Paul., 10: 728 (Piracicaba;
Rio Verde, Minas; Itatiba; Rio Mogy-Guassu, Salto do Pirassununga); Devincenzi,
1924, An. Mus. Nac. Montevideo, (2) 1: 152 (Rio Uruguay, frente al Salto) ; Devin-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

199

cenzi & Teague, 1942, An. Mus. Hist. Nat. Montevideo, (2) 5: 42 (rio Uruguay, frente
a Concordia); van der Stigchel, 1946, South American Nematognathi: 44 (Parma River,
Rosario de Santa Fe); van der Stigchel, 1947, Zool. Meded., 27: 44 (Parana River,
Rosario de Santa Fe).

Zungaro mangurus; Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 384
(La Plata); Bertoni, 1914, Fauna Paraguaya: 8 (Alto Parana); Gomes, 1946, Occ. Pap.
Michigan Mus., 494: 6 (eaux douces du Bresil et du Paraguay).

Pseudopimelodus parahybae ; Devincenzi & Barattini, 1926-1940, An. Mus. Hist. Nat.
Montevideo, Supl.: pi. 52 (Uruguay); Devincenzi & Teague, 1942, An. Mus. Hist. Nat.
Montevideo, (2) 5: 42 (Uruguay).

Pseudo pimelodus variolosus\ Pearson, 1937, Proc. Calif. Acad. Sci., (4) 23: in (Para¬
guay basin).

Cephalosilurus mangurus ; Gosline, 1941, Stanford Ichth. Bull., 2: 85 (Brazil and
Paraguay); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 27 (Brasil e
Paraguai).

Pseudo pimelodus roosevelti ; Gosline, 1941, Stanford Ichth. Bull., 2: 85 (Parassununga,
State of Sao Paulo); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 28 (Sao
Paulo) ; (pt.) Gomes, 1946, Occ. Pap. Michigan Mus., 494: 4, 7 (Rio Piracicaba, Piraci-
caba, Sao Paulo) ; Gomes, 1956, Rev. Bras. Biol., 16: 404 (no locality); Schubart, 1964,
Bol. Mus. Nac. Rio de Janeiro, Zool., 244: 18 (Rio Mogi Gua^u); Travassos & Kohn,
1965, Pap. Avuls. Zool. S. Paulo, 17: 44, 45 (Pirassununga); Lundberg & Baskin, 1969,
Amer. Mus. Novit., 2396: 6, 29 (no locality).

Zungaro zungaro ; Fowler, 1943, Proc. Acad. Nat. Sci. Philad., 95: 318 (Arroyo
Canelones; Rio Santa Lucia, Canelones) ; de Buen, 1950, Publ. Cient. S.O.Y.P., 2: 71
(Uruguay) ; Bertin & Esteve, 1950, Cat. Types Mus. Paris, 5: 55 (Rio de la Plata) ;
Ringuelet, Aramburu & de Aramburu, 1967, Los Peces Argentinos de Agua Dulce: 325
(Argentina: Rio Parana medio e inferior; Rio Paraguay; Rio Uruguay media e inferior;
Rio de la Plata).

Microglanis cottoides; De Buen, 1950, Publ. Cient. S.O.Y.P., 2: 70 (Uruguay).

Material. — One specimen, December 1907, Parassununga, Sao Paulo, Brazil (R. von
Ihering, AMNH no. 26942, formerly no. 8638a), standard length 140 mm, paratype of
Pseudo pimelodus roosevelti . One specimen, 9 September 1908, Piracicaba, Rio Tiete, Sao
Paulo (J. D. Haseman, FM no. 58120), standard length 107 mm. Two specimens, 15 Sep¬
tember 1908, Salto de Avanhandava, Rio Tiete, Sao Paulo (J. D. Haseman, FM
no. 58070), standard length 52, 56 mm. Two specimens, same data (FM no. 58122),
standard length 21, 33 mm. One specimen, 7 March 1925, Rosario de Santa Fe, Parana,
Argentina (Noordraven, ZMA no. 100138), standard length 195 mm. 13 specimens, 25
March 1945, Cachoeira de Emas, Rio Mogi Guassu, Est. de Sao Paulo (G. S. Myers,
SU no. 62437), standard length of eight specimens 26-35 mm > five specimens 108-151 mm.

Diagnostic characters. — Very close to P. z. bufonius, but colour pattern
slightly different: the cross-bands of the body are not connected dorsally,
or only just connected, the second dark band is usually not continued down¬
wards on to the anal fin (but in small specimens it is), arid the dark cross-bar
on the caudal fin is broader, and almost terminal, leaving only a narrow white
edge, broadest at the tip of the upper lobe.

Description. — D 1 . 6 , A 11-12 (iv or v. 6 or 7), P 1.6 or I.7, V 6 (1.5),
C i.i4.i or i.is.i and rudiments, gill-rakers 1 + 1 + 2 to 5; branchiostegals
8 or 9: teeth in both jaws in a band, the band of premaxillary teeth with

200

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

acute projections laterally, which, however, do not reach far backwards (fig.
37a); eyes small, their greatest diameter 3.0-3.5 times in snout, 2.5-4.0 times
in bony interorbital, 4.3-5.5 times in whole interorbital, and 9.5-11.0 times
in head; anterior nostrils well behind upper lip; the maxillary barbels reach
just to the pectoral base; the outer pair of mental baihels is a little more than
half their length, and the inner pair is about half the length of the outer pair.

Head 3.0-3.7 times in standard length, depressed, its depth about 0.6 times
its width, with a narrow occipital process meeting the dorsal plate; greatest
width of body, between the cleithra, 3.35-4.0 times in standard length;
greatest depth of body 4.8-5.0 times in standard length; predorsal length
2.6-3.05 times in standard length.

Dorsal fin with one spine and six divided rays, the spine strong, but
short, with a blunt tip, its length just equal to snout and eye and 2.6-3.S
times in predorsal length, along both edges serrations, which become visible
only after removing the skin; soft rays about one and a half times the length
of the spine, which is still comparatively short, when pressed down they reach
only from one third to half way the distance to the adipose fin. Anal fin
rounded, with four or five simple and six or seven divided rays, of the same
length as the dorsal rays, the longest rays a little over half the length of the
head. Pectorals with one spine and seven rays; the spine is strong, with along
each edge some 18 teeth, those on the outer edge small, and only slightly
directed inwards, those on the inner (posterior) edge larger, though less
than half the width of the spine and distinctly directed inwards; the spine is
encased in thick leathery skin which is continued beyond its tip; the whole
spine is strongly curved backwards; without removing the skin, the teeth
along the anterior edge cannot be seen, the teeth along the posterior edge are
continued along the leathery part as in P. z. bufonius\ the combined length
of the pectoral spine and its leathery tip equals the length of the head, minus
the distance from tip of snout to posterior nostril, and is therefore a little
longer, relatively, than in P. z. bufonius ; the first ray is equal in length to
spine and tip, but the last one is no more than half the length of the first
one, and equals snout with eye. Ventrals rounded, with six rays, of which
the first one is simple, the others divided, reaching two-thirds of the distance
to the anal fin, the longest rays equal in length to the length of the longest
anal rays. Caudal fin well-developed, forked with acute tips, with 14 divided
rays, above and below one fully developed simple ray, and rudiments. Adipose
fin of moderate length, its origin opposite the origin of the anal fin and its
base reaching very slightly farther than the base of the anal fin, its height
one third of the depth of the caudal peduncle below it.

Colour of preserved specimens. Dark brown bands and patches on a pale

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

201

brownish background. The head is either pale, with scattered small dark
spots, not forming a pattern, or entirely dark; there is a dark band across the
nape, reaching to the lateral line; a second, much wider dark band from
the dorsal fin downwards and backwards, reaching to a little below the lateral
line (in Borodin’s figure of the type of P. roosevelti, the first band is barely
indicated, the second reaches to the lateral line; in the examined paratype the
second band reaches to below the lateral line, but not so far down as in P.
z. bufonius); a third dark area is on the back, anteriorly not or just connected
with the second one, posteriorly taking in most of the adipose fin and reaching
down about two-thirds of the body-depth to above the origin of the anal fin;
a round pale patch on the back just in front of the adipose fin is excluded
from this dark area; a complete dark band, of equal width over its length,
over the distal part of the caudal peduncle and the base of the caudal fin.
The dorsa'l fin is mostly dark, with a pale spine, pale tips to the rays, and a
narrow pale median cross-bar over the rays. The anal fin is pale with two
broad dark cross-bars. The pectorals are pale, but the rays (not the spine)
have dark bases, and there is a broad but ill-defined dark band over the
middle of the fin. The ventrals are pale, with a dark base and a dark median
crossbar. The tail has a dark base, as already mentioned, followed by a
broad pale band, which is followed by a subterminal dark band of about the
same width as the white band; there is a narrow pale margin, and often a
somewhat longer pale tip to the upper caudal lobe. The figures published by
Borodin (1927), under the name Pseudopimelodus roosevelti , by Devincenzi
& Barattini (1926-1940: pi. 52), and Devincenzi & Teague (1942: 42), under
the name Pseudo pimelodus parahybae, show the characters of P. zungaro man-
gurus very well.

Distribution. — Probably throughout the Uruguay-Parana-Paraguay-La
Plata basins.

Discussion. — The figure on which this subspecies was founded (cf. Valen¬
ciennes, 1847) unsatisfactory, but d’Orbigny forwarded also a mounted
specimen, which was re-identified as Zungaro zungaro by Bertin & Esteve
(1950), and therefore undoubtedly belongs to the present species.

As mentioned above, the fish described and figured by Devincenzi &
Barattini (1926-1940: pi. 52) and Devincenzi & Teague (1942: 42) under
the name Pseudo pimelodus parahybae , clearly belongs to this species. Note
that the mentioned authors record a length of 40 to 50 cm and a weight
of 5 to 6 kg for their specimens, which is quite impossible for Microglanis
parahybae. Devincenzi & Teague (1942) distinguish between Pseudopimelo-
dus parahybae and Pseudo pimelodus zungaro, but, although the mentioned
authors do not express themselves very clearly on the subject, one gets an

202

ZOOLOGISCHE VERHANDELINGEN 132 ( 1974)

impression that they had not examined a specimen which they could identify
with the second species, and included it on the testimony of a former prepa-
rator of their museum.

Pseudopimelodus zungaro zungaro (Humboldt( (fig. 39, pi. 5)

Pimelodus zungaro Humboldt, 1833, Rec. Observ. Zool. Anat. Comp., 2: 170, pi. XLVI
fig. 1 — Tomependa, sur les bords de TAmazone.

Zungaro Humboldtii Bleeker, 1858, Visschen Ind. Arch., I : 206 — nomen novum for
Pimelodus zungaro Humboldt, to avoid tautonomy.

Pimelodus (Pseudopimelodus) pulcher Boulenger, 1887, Proc. Zool. Soc. Lond.: 276,
pi. XXI fig. 1 — Canelos.

Pseudopimelodus variolosus A. de Miranda Ribeiro, 1914, Comm. Linh. Tel. Estr.
Matto-Grosso ao Amazonas, 15, Annexo 5: 4, pi. 1 fig. 2, pi. 2 fig. 1,2 — Coxim, no
Rio Taquary, M. Grosso; S. Manoel.

Pimelodus Zungaro; Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.), 15: 119 (le haut Maragnon, dans la province de Jaen Bracamoros); R. H.
Schomburgk, 1843, Fish. Brit. Guiana, 2: 205 (the Amazon near Tomependa).

Zungaro Humboldtii ; Bleeker, 1862, Atlas Ichth., 2: 11 (name only); Bleeker, 1863,
Ned. Tijdschr. Dierk., 1 : 101 (name only).

Pimelodus bufonius ; Cope, 1878, Proc. Amer. Philos. Soc., 17: 675 (Peruvian Amazon).

Pseudopimelodus pulcher ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) 1: 122 (no locality); Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci.,
1: hi (Eastern Ecuador); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14:
27 (Eastern Ecuador); Fisher, 1917, Ann. Carnegie Mus., n : 410, fig. 1 (Salto Avan-
handava, Rio Tiete); Borodin, 1927, Amer. Mus. Novit., 266: 3 (Ecuador); Pearson,
1937, Proc. Calif. Acad. Sci., (4) 23: 92 (Tingo de Pauca, Rio Maranon, Peru) ; Fowler,
1941, Bol. Mus. “Javier Prado", 5: 378 (Peru (Rio Maranon). Tambien en el Alto
Amazonas en el este del Ecuador); Eigenmann & Allen, 1942, Fish. W. South America:
91 (Eastern Ecuador and Upper Maranon); Fowler, 1945, Los Peces del Peru: 47
(Peru (Rio Maranon)'. Tambien en el Alto Amazonas en al este del Ecuador) ; Gomes,
1946, Occ. Pap. Mus. Michigan, 494: 6 (Tingo de Pauca, at the mouth of Rio Crisnejas
in Rio Maranon, Peru); Fowler, 1951, Arq. Zool. S. Paulo, 6: 561 (Alto Amazonas,
Equador, Peru, Sao Paulo) ; Ovchynnyk, 1968, Zool. Anz., 181: 254 (Ecuador: Canelos
on Rio Bobonaza, tributary of Rio Pastaza, Prov. Napo-Pastaza).

Batrachoglanis pulcher ; Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3:
383 (Eastern Ecuador).

Zungaro zungaro ; Fowler, 1915, Proc. Acad. Nat. Sci. Philad., 67: 209 (Peruvian
Amazon); Fowler, 1940, Proc. Acad. Nat. Sci. Philad., 91 : 285 (Peruvian Amazon) ;
Fowler, 1940, Proc. Acad, Nat. Sci. Philad., 92: 95 (Bolivia: Huachi; Popoi R., upper
Beni; R. Colorado, lower Bopi); Fowler, 1941, Bol. Mus. “Javier Prado", 5: 378,
fig. 17 (Peru : Rio Maranon, Amazonas peruano, figured specimen from Rio Ampiyaco) ;
(pt.) Eigenmann 8 c Allen, 1942, Fish. W. South America: 90 (throughout upper Ama¬
zon) ; (pt.) Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 33 (Maranon) ;
Fowler, 1945, Los Peces del Peru: 48, fig. 17 (Peru: Rio Maranon, Amazonas peruano);
P. de Miranda Ribeiro, 1964, Bol. Mus. Nac. Rio de Janeiro, Zool., 246: 3 (Rio
Araguaya).

Pseudopimelodus zungaro ; Pearson, 1924, Indiana Univ. Stud., n (64): 10 (Bolivia:
Huachi and Popoi rivers, Upper Beni; Rio Colorado, Lower Bopi); Pearson, 1937,
Proc. Calif. Acad. Sci., (4) 23: in (Beni basin).

Pseudopimelodus variolosus ; Gosline, 1941, Stanford Ichth. Bull., 2: 85 (Rio Taquary,
State of Matto Grosso, Brazil) ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool,

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

203

33: 28 (Mato Grosso); Gomes, 1946, Occ. Pap. Michigan Mus., 494: 7 (Rio Taquiri, Mato
Grosso, Brazil) ; (pt.) Fowler, 1951, Arq. Zool. S. Paulo, 6: 562, fig. 570 (Rio Taquari);
(pt.) P. de Miranda Ribeiro, 1953, Arq. Mus. Nac. Rio de Janeiro, 42: 404 (Coxim,
no Rio Taquari, Mato Grosso; Sao Manoel).

Microglanis pulcher ; Gosline, 1941, Stanford Ichth. Bull., 2: 85 (Canelos, Ecuador);
Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 28 (este do Equador).

Material. — One specimen, “1883”, Canelos, Ecuador (purchased from Gerrard, MV
no. 45616), standard length 63 mm 1 ). Three specimens, September 1921, Huachi, Popoi
River, upper Beni, Bolivia (N. E. Pearson, CAS no. 17954), standard length 96, 98,
140 mm. One specimen, September 1923, Tingo de Pauca, Maranon River, Peru (N. E.
Pearson, CAS no. 17923), standard length 87 mm. One specimen, 2 May 1968, Sangadina,
Mato Grosso (R. H. McConnell, coll. no. 125A), standard length 56 mm.

Fig. 39. Dentition and right-hand pectoral-spine of “Pseudopimelodus pulcher ”

(= P. z. sungaro).

Characters. — Similar to P. z. mangurus, but band over the caudal fin
narrow, except in a few specimens; similar to P . z. bufonius, but dark bands
on body not dorsally connected .

Discussion. — Originally I believed that P. pulcher was a distinct species
as the only specimen then available (MV no. 45616) was not only charac-

1) It is practically certain that this specimen was collected by C. Buckley in 1880 and
belonged to the same lot as the types of P. pulcher , for Boulenger (1887: 274) mentioned
that part of Buckley's collection was: “Sold by the well-known dealer Mr. Gerrard to
other institutions, principally to the Vienna Museum”. It bore the identification P.
bufonius .

204

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

terized by its remarkable tail-pattern, entirely agreeing with Boulenger’s
(1887) figure and different from any specimen of P. zungaro I had seen
at the time, but also appeared to differ in certain fin-ray and gill-raker
numbers. It had D 1 . 6 , A io (iv.6), P L6, V 6 (i.5), C i.i3.i and rudiments,
gill-rakers 2+1 + 5.

The first doubt arose when I examined sample FM no. 58122, from a
locality remote from the type-locality of P. pulcher . This sample had been
identified as P. pulcher by Fisher (1917), an identification changed to P.
roosevelti by Gomes (1946). Nevertheless, as Fisher mentioned, the larger
of the two specimens agrees very well with P. pulcher, in tail pattern, pec¬
toral fin (P 1 . 6 ), and gill-rakers (1 + 1 + 5).

Later I found specimens with a caudal-fin pattern intermediate between
that of typical P. pulcher and the ordinary cross-bar of P. zungaro. Further
it became evident that in small specimens of P. zungaro the number of
pectoral fin rays is quite often six instead of the more usual seven. One
specimen identified as P. pulcher (CAS no. 17973) on the left P 1 .7, on
the right P 1 . 6 . As regards the gill-rakers, apparently small specimens tend
to have one or two more (usually very rudimentary ones) than larger indivi¬
duals.

There can be no reasonable doubt that P . variolosus is a synonym. In its
description A. de Miranda Ribeiro (1914) states D I + 5, A 7, but his figure
shows D 1 . 6 , and as far as the anal fin is concerned, it may be assumed that
he has counted the divided rays only and not the three or four simple rays
which precede them. In other respects, description and figure agree well with
P. zungaro .

Pseudopimelodus zungaro subsp . 1 )

Pseudopimelodus bufonius; (pt.) Steindachner, 1880, Denkschr. Akad. Wiss. Wien,
Mathem.-Naturw. Cl., 42 (1) : 59 (Cauca) ; Miles, 1947, Peces Rio Magdalena: 58, 64
(sistema Magdalena-Cauca).

Pseudopimelodus zungaro; Eigenmann, 1922, Mem. Carnegie Mus., 9: 32 (Quibo, Rio
Atrato; Soplaviento, El Digue; Honda, R. Magdalena; Puerto Berrio, R. Magdalena;
El Banco, R. Magdalena); Miles, 1943, Peces de Agua Duke del Valle del Cauca
(Cauca), non vidi.

Material. One specimen, 11/13 January 1912, Soplaviento (C. H. Eigenmann, FM
no. 57699), standard length 171 mm. One specimen, 21 January 1912, El Banco (C. H.
Eigenmann, FM no. 57700), standard length 207 mm. Two specimens, January 1912,
El Banco and Soplaviento (C. H. Eigenmann, CAS no. 17975, ex IUM nos. 13498,

1) Perhaps the name Zungaro zungaro schultzi Dahl 1955 is applicable to this sub¬
species. I have been unable to consult the original description, but found the name
listed as a synonym of Pseudopimelodus bufonius in Dahl (1971: 54). The fish figured
by Dahl under the name P. bufonius shows a long, rounded-lanceolate caudal fin, quite
unlike any race of P. zungaro known to me, with a narrow dark cross-band.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

205

13499), standard length 112, 149 mm. One specimen, 10 June 1956, Rioviego, R. Magda¬
lena, Bolivar (G. Dahl, USNM no. 175310), standard length 101 mm. One specimen,
no date, Honda (Niceforo Maria, USNM no. 121258), standard length 151 mm.

Characters. — Very similar to P. z. bufonius, and mainly distinguishable
by the colour-pattern of the caudal fin, which is entirely dark, except for a
rather narrow pale cross-bar near its base, and white tips. Hence, the tail
is even darker than in the nominate race. It looks as if the amount of white
on the anal fin is also reduced, the white median band being less developed,
and that also the pale caudal pedunde is less developed, but all specimens
examined were more or less decolourized. Number of gill-rakers 1 + 1+4
(four times) and 1 + 1 + 3 (twice).

Distribution. — Only known from the Rio Magdalena basin, Colombia.

Discussion. — The first to record this species from Colombia was Stein-
dachner (1880), who had a specimen of 55 cm length from Cauca. Note,
however, that the excellent figure (pi. II fig. 1) which accompanies his
paper, and which shows a dark caudal band of moderate width, cannot have
been taken from the Colombian specimen as its figured total length, at % X
natural size, is a little less than 22 cm, hence its actual length must have been
ca. 29 cm.

There has been the usual confusion about the identity and nomendature:
Eigenmann (1922) used the name Pseudopimelodus zungaro for material
from the Rio Magdalena, but note that he (Eigenmann, 1922: 32) was not
very sure, as he mentioned in the discussion of P. zungaro that: “the large
specimens reported may belong to another species”. Miles (1947) rejected
Eigenmann’s identification and used the name P. bufonius for specimens
from the Magdalena and Cauca, as Steindachner (1880) had done previously.

Pseudopimelodus raninus (Valenciennes)

Under this name I unite four forms which, although they have hitherto
been regarded as distinct species, are so similar that in my opinion there is
every reason to treat them as subspecies of one species. There are scarcely
any morphological differences between the four forms (one is a little more
slender than the others), but they differ in colour-pattern, a difference that,
at least when specimens of similar sizes are compared, appears to be constant.
Just as in P. zungaro, and for the same reasons, I have given a fairly exten¬
sive description of the subspecies occurring in Suriname, which happens to
be the nominate one, and only diagnoses indicating their differential charac¬
ters, of the others.

This species is not known to attain a large size; the largest specimen

206

ZOOLOGISCHE VERHANDELINGEN 1 32 (1974)

examined by me had a standard length of 166 mm, and Schultz (1944a)
mentioned one of 184 mm.

The distribution includes the whole Amazon basin, French Guiana and
Suriname (P. r. raninus); the Essequibo and Orinoco basins (P. r. villosus );
the Maracaibo basin (P. r. acanthochiroides), and rivers west of the Andes
in Colombia and Ecuador (P. r. transmontanus).

P. raninus has also been recorded from the La Plata basin, in Paraguay
and Argentina (Pozzi, 1945, non vidi, and Castello, 1971), but at least one
of these records was based on a misidentification, and the occurrence of
R. raninus in the La Plata basin requires confirmation 1).

Pseudopimelodus raninus raninus (Valenciennes) (fig. 37b, 38b, pi. 8, 9)

Pimelodus raninus Valenciennes, 1840, in Cuvier & Valenciennes, Hist. Nat. Poiss.
(4 0 ed.),i5: 117, pi. 434 — La Mana.

Pseudopimelodus acanthochirus Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) 1: 122 — Gurupa, Tajapuru, Teffe, Jutahy.

Pimelodus raninus Kner, 1858, Sitzb. Akad. Wiss. Wien, Mathem.-Naturw. Cl., 26
(1857): 421 (Barra do Rio negro, Guapore und Matogrosso); (?) Regan, 1905, Proc.
Zool. Soc. Lond.: 190 (Rio Negro).

Pseudopimelodus raninus ; Bleeker, 1858, Visschen Ind. Arch., 1: 207 (Amer. merid.) ;
Bleeker, 1862, Atlas Ichth., 2: 11 (name only); Bleeker, 1863, Ned. Tijdschr. Dierk., 1:
101 (name only) ; Steindachner, 1880, Denkschr. Akad. Wiss. Wien, Mathem.-Naturw.
Cl., 42 (1) : 61 (no locality); Steindachner, 1883, Denkschr. Akad. Wiss. Wien, Mathem.-
Naturw. Cl., 56 (1) : 4 (Rio Huellaga) ; Eigenmann & Eigenmann, 1888, Proc. Calif.
Acad. Sci., (2) 1: 122 (name only) ; (pt.) Eigenmann & Eigenmann, 1890, Occ. Pap.
Calif. Acad. Sci., 1: in (reference to original description only); (pt.) Eigenmann &
Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 27 (compiled); Eigenmann, 1912, Mem.
Carnegie Mus., 5: 154 (la Mana); A. de Miranda Ribeiro, 1914, Comm. Linh. Telegr.
Estr. Matto Grosso ao Amazonas, 15, Annexo 5: 3 (Mantaos); (pt.) Eigenmann & Allen,
1942, Fish. W. South America: 91 (compiled); Gosline, 1945, Bol. Mus. Nac. Rio de
Janeiro, Zool., 33: 28 (Guiana Francesca?); Gomes, 1946, Occ. Pap. Mus. Michigan, 494:
7 (“Mana” ? French Guiana); Hoedeman, 1961, Bull. Aquatic Biol., 2: 135 (Aloike,
Litany river, “French Guiana”); Marlier, 1968, Cad. Amazonia, 11: 53 (le lac du Rio
Preto da Eva).

1) Dr. Castello has forwarded two of the specimens on which his record was based;
they are from Vicente Lopez near Buenos Ayres (MACN no. 3373)- Their dentition
indicates that these specimens are referable to Microglanis and not to Pseudopimelodus :
the band of teeth in the upper jaw is laterally rounded, without any trace of a backward
projection; the lateral line has pores to a little behind the dorsal base. The large specimen
measures 54 mm in standard length, D 1 . 6 , A 13 (v.8), P 1 . 6 , V 6 (i.5), C 13 (branched
rays only), branchiostegals 9, gill-rakers 2 + 1+8, head 3.1, greatest width of body 3.0,
greatest depth of body 4.3, predorsal length 2.6 times in standard length. The other
specimen measures 49 mm in standard length, D I.7, An (iv.7), P 1 . 6 , V 6 (i.5), C 13
(branched rays only), branchiostegals 9, gill-rakers (left) 2+1 + 6, (right) 2+1 + 5,
head 3.0, greatest width of body 3.0, greatest depth of body 4.5, predorsal length 2.5 times
in standard length. I believe that these specimens fall within the range of variation of
Microglanis parahybae , a very variable species. What remains of the colour-pattern sup¬
ports this identification.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

207

Pseudopimelodus acanthochira; Eigenmann & Eigenmartn, 1890, Occ. Pap. Calif.
Acad. Sci., 1: 114 (Amazon from Para to Tabatinga: Gurupa, Tajapura, Teffe, Jutahy);
Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 28 (Amazon, Solimoens) ;
Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3: 384 (Amazonas from Para
to Tabatinga); A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro, 16: 257
(no locality); Giintert, 1942, Zool. Anz., 138: 31 (no locality); Gosline, 1945, Bol. Mus.
Nac. Rio de Janeiro, Zool., 33: 28 (Baixo Amazonas).

Batrachoglanis rartinus ; Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3:
383 (Rio de Janeiro to Essequibo and Peru).

Pseudopimelodus villosus ; (pt.) Fisher, 1917, Ann. Carnegie Mus., 11: 410 (Santarem:
the two larger specimens).

Pseudopimelodus acanthocheira ; Fisher, 1917, Ann. Carnegie Mus., 11: 411, fig. 2
(Maciel, Rio Guapore); Pearson, 1937, Proc. Calif. Acad. Sci., (4) 23: in (Mamore
basin).

Pseudopimelodus acanthochirus ; Gomes, 1946, Occ. Pap. Michigan Mus., 494: 6
(Gurupa and Tajapuru, Para; Tefe and Jutai, Amazonas).

Zungaro raninus ; Bertin & Esteve, 1950, Cat. Types Poiss. Mus. Paris, 5: 55 (FI.
La Mana (Guyane)).

Microglanis parahybae ; Emmens & Axelrod, 1968, Catfish: 28, pi. (no locality ) 1 ).

Material. — 45 specimens from Suriname as follows. Five specimens, 27 November
1957, “Aloike”, Litani Rivier (J. R. Gery, ZMA no. 102232), standard length 48-82 mm.
One specimen, 19 December 1965, Compagnie Kreek (Mees, RMNH no. 26602), standard
length 49 mm. Three specimens, 28 December 1965, Marshall Kreek (Mees, RMNH
no. 26603), standard length 45, 69, 79 mm. One specimen, 20 October 1966, Gran Kreek,
ca. 63 km S. of Affobakka (Nijssen, ZMA no. 105947), standard length 61 mm. One
specimen, 8 December 1966, Marshall Kreek (Nijssen, ZMA no. 106371), standard length
43 mm. 34 specimens, 21 April 1967, Maka-kreek, a small tributary of the Lawa Rivier,
ca. io km S. of Stoelmanseiland (Nijssen, ZMA no. 105950), standard length 19-75 mm.

Extra-limital material. — One specimen, before 1857, Barra do Rio negro (no collector,
MV no. 44439), standard length 88 mm. Two specimens, undated, Tajapuru (no collector,
presumably Thayer Exp., MCZ no. 9833), standard length 89, 93 mm. One specimen,
before 1883, Huallaga, Peru (A. Stiibel, MTD no. 1649), damaged, see discussion. Ten
specimens, July-August 1909, Maciel, Rio Guapore, Brazil (J. D. Haseman, FM
no. 58121), standard length 25-41 mm. Two specimens, 15-19 December 1909, Santarem,
Brazil (J. D. Haseman, FM no. 58140), standard length 34, 120 mm. Two specimens,
undated, Boby Creek, Approuage R., French Guiana (C. Ternetz, BM 1926-3-2: 703-704),
standard length 41, 68 mm. One specimen, July 1965, Rio Aspeie, Boa Vista, Municipio
of Castanhal, Para, Brazil (N. Menezes, MCZ no. 46031), standard length 36 mm. Two
specimens, July 1965, Rio Arari, Cachoeira do Arari, Ilha de Marajo, Pana, Brazil (N.
Menezes, MCZ no. 46033), standard length 33, 34V2 mm. 14 specimens, 24 April 1967,
Kamaloea- or Saloea Kreek, a tributary of the Marowijne, 9 km S.E. of the mouth of
the Gran Kreek, French Guiana (Nijssen, ZMA no. 105952), standard length 28-78 mm.

Diagnostic characters. — Head in size between that of P. nigricauda and
P. z. bufonius, slightly angular at the gape; head and body dark above, with
a fairly narrow but conspicuous pale band across the nape, reaching from
pectoral to pectoral, and another pale band on the caudal peduncle, including

1) The fish described in the text of this book appears actually to be Microglanis
parahybae, but the coloured photograph on page 28 is clearly of P. raninus raninus , and
I have listed it as it is an excellent illustration of this form.

208

ZOOLOGISCHE VERHANDELINGEN I32 (I974)

the posterior half of the adipose fin; pectoral fin with six rays; maxillary
barbels reaching to or almost to a vertical through the origin of the dorsal
fin; the dorsal fin when depressed reaches from two-thirds to nearly to the
adipose fin; caudal fin scalloped, with two more or less equal lobes, hyaline
white with a black base, a black band across, and a number of brown dots
in between; the short occipital process narrowly separated from the much
longer dorsal plate in small specimens, touching it in larger individuals; gill-
rakers well-developed, slender.

Description. — D 1 . 6 , A io-ii (iii.7, iii.8, or ii.9, usually 8 divided rays),
P 1 . 6 , V 6 (i.5), C with 13 or 14 branched rays; gill-rakers well-dieveloped,
slender, 3 4- 1 + 7-10; branchiostegals 9; teeth in both jaws in a broad
band, the band of premaxillary teeth laterally projecting backwards; eyes
small, their greatest diameter 2.5-3.0 times in snout, about 4 times in inter-
orbital, and 9-11 times in head; the maxillary barbels are relatively long,
reaching to well beyond the basis of the pectoral fin, and often to the vertical
through the origin of the dorsal fin; the outer pair of mental barbels reaches
to the basis of the pectoral fin, or slightly beyond, the inner pair is about
two-thirds the length of the outer pair.

Head 2.9-3.0 times in standard length, greatest width of body 3.2-3-3 times
in standard length, greatest depth of body 3.8-44 times in standard length,
predorsal length 24-2.5 times in standard length.

Dorsal fin with one spine and six divided rays; dorsal spine moderate,
equal to snout with eye or a little longer, encased in skin, without serrations,
or these very weakly indicated; soft rays moderately long, when depressed
reaching from two-thirds to just to the adipose fin, longest dorsal ray equal
to postorbital part of head or slightly longer. Anal fin with three simple and
seven to nine (usually eight) divided rays, rounded, its longest ray about half
the length of the head. Pectorals with one spine and six divided rays; the
spine is veiy strong, flat, curved slightly backwards, varying from somewhat
less to well over half the length of the head, its anterior and posterior borders
each with a series of from twelve to fifteen strong spines; the first ray is
longer than the spine, the following rays are successively shorter, and the
last ray is half to two-thirds the length of the first ray. Ventrals rounded,
with six rays, the first one simple, the others divided, relatively long, reaching
almost to the anal fin, and the longest rays equal to the distance from the
gape to the pectoral fin or even slightly longer. Caudal fin with two rounded
lobes, the upper lobe usually slightly the longer, with thirteen or fourteen
divided rays, and above and below several simple ones: dorsally as many as
eight, definitely influencing the shape of the fin; the longest ray equal to
head without half the snout. Adipose fin moderate, implanted opposite the

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

209

anal fin, and its base of the same length as that of the anal fin, its height
about two-fifths the depth of the caudal peduncle.

Colour. Generally blackish brown, underparts from chin to anal fin dull
brown, the pigment spots not covering the whole skin. From pectoral to pec¬
toral, across the nape, is an irregular pale bar. A pale brown band across the
posterior part of the body, from adipose to anal fin; in small specimens
sometimes constricted or even interrupted on the lateral line. A small pale
saddle just behind this band. Barbels ringed blackish-brown and white. The
dorsal fin is black, with a white edge, and with over the last four rays and
their connecting membrane near their base a white band; the anal fin is black
With a white edge, and with on the basal half in its posterior part, over the
last six rays, a white band; the pectoral fins are 'black, with the basal third
of the spine pale, and a small white band, sometimes almost obsolete, over
the last four rays near their bases; caudal fin with a blackish brown basis
and a blackish brown subterminal band, the white area in between with light
brown freckles, variable in intensity and extent; the ventrals are black, with
on the last two or three rays near their base usually indications of a white bar;
adipose fin with on its anterior base the pale spot already mentioned, middle
section black, posterior half pale brown.

The preceding description was based on material from Suriname, and small
specimens from the Lower Amazon appear to agree entirely, in morphology
as well as in colour-pattern. Specimens from the Rio Guapore (FM no.
58121) do not differ, except that some have a pale blotch of irregular shape
laterally just behind the dorsal fin.

A large Amazonian specimen (FM no. 58140, measuring 120 mm) is very
different; with its spotted appearance it greatly resembles the next sub¬
species. Amazonian specimens of intermediate length are old and faded, but
may have been similar.

The question whether large specimens from Suriname would also lose their
distinctive pattern, even the question whether specimens from Suriname do
ever attain such a large size, remains to be answered. See also the discussion
of F. r. villosus .

Distribution. — Suriname and French Guiana: Suriname, Marowijne,
Mana and Approuague basins. It may be significant that in Suriname this
subspecies has not been found further west than the Suriname Rivier basin;
in the Demerara River and the Essequibo it is replaced by a different sub¬
species. Brazil: widely distributed in the Amazon basin, upstream into Peru
and to the Bolivian border. Specimens from the Paraguay and the La Plata
rivers may also belong to this subspecies. Records from Rio de Janeiro are
erroneous, see discussion.

14

210

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Discussion. — There has been some confusion about the type-locality of
this species; some authors have sought la Mana in the neighbourhood of Rio
de Janeiro, and others (Gosline, 1941, 1945; Gomes, 1946) marked the
locality French Guiana with a query. This uncertainty was caused by careless
reading of Valenciennes (1840) who had to say this about the provenance of
his material: “II a ete envoye de la Mana au Cabinet du Roi par MM.
Leschenault et Doumerc en 1824, et nous voyons par un dessin de M. Mene-
strier, qu’il se porte au sud au moins jusqu'a Rio-Janeiro”. This shows con¬
clusively that Valenciennes's material came from the Mana, and that Rio de
Janeiro was mentioned on the basis of a plate. Nowhere does Valenciennes
say that the Mana is anywhere near Rio de Janeiro, and the fact that
Leschenault visited the Mana River in French Guiana in 1824 (van Steenis-
Kruseman, 1950), shows that this and no other river is meant. Eigenmann &
Eigenmann (1890) indicated it correctely on their map. The plate from
Rio de Janeiro would doubtless pertain to a different species, and as the only
actual specimens are from the Mana, they may be regarded as the types; see
also Bertin & Esteve (1950).

In the older literature (Muller & Troschel in Schomburgk, 1848; Kner,
1858; Gunther, 1864; Peters, 1877; Steindachner, 1883), Pseudopitnelodus
raninus has been recorded from a series of widely scattered localities in
South America. At the time it was not known that Pseudo pitnelodus contains
a number of superficially often very similar species and subspecies, several
of which appear to have a restricted distribution. As mentioned on a previous
page, Eigenmann (1912) observed that Gunther's specimens do not belong to
P. raninus, and also that even at that time the material recorded by Muller &
Troschel and by Peters was in such a bad state of preservation that identifi¬
cation was no longer possible. Schultz (1944a: 198) referred to a specimen
from( the Rio Meta in Colombia. He failed to give information on the
depository of this specimen so that I cannot check its identity, but it seems a
safe assumption that it was misidentified and belonged to the following sub¬
species. I have examined the specimen from the Rio Huallaga, Peru (leg.
Stiibel, MTD no. 1649), recorded by Steindachner (1883: 4). Unfortunately
its liquid has 'boiled away in the fire following a bomber-attack in 1945, and
all that is left are a few charred remains, the most important ones of which
are the pectoral girdle with both spines, and a piece of the cranium; the anal
fin appears to have had ten rays. These remains are scarcely identifiable:
until such time as fresh material from the Rio Huallaga may become available
for comparison, I cannot definitely state that they belong to P. r. raninus,
but they do certainly not contradict Steindachners's identification, which I
believe to have been correct.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

211

Pseudopimelodus raninus villosus Eigenmann

Pseudopimelodus villosus Eigenmann, 1912, Mem. Carnegie Mus., 5: 152, fig. 32;
pi. X fig. 1 —Potaro Landing, British Guiana.

Pseudopimelodus villosus Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3:
384 — nomen nudum (British Guiana).

Pimelodus raninus ; Gunther, 1864, Cat. Fish. Brit. Mus., 5: 133 (Essequibo).

Pseudopimelodus villosus; Caporiacco, 1935, Monit. Zool. Ital., 46: 58 (Curupucari
(Essequibo)); Gosline, 1941, Stanford Ichth. Bull., 2: 85 (British Guiana); Gosline,
1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 28 (Guiana Inglesa); Gomes, 1946, Occ.
Pap. Michigan Mus., 494: 2 (no locality); Lowe, 1964, J. Linn. Soc. Lond., Zool., 45: 140
(Rupununi); Alexander, 1964, Proc. Zool. Soc. Lond., 142: 420 (British Guiana);
Alexander, 1966, Proc. Zool. Soc. Lond., 148: 90 (British Guiana).

Pseudopimelodus villosus villosus ; Schultz, 1944, Proc. U.S. Nat. Mus., 94: 198
(British Guiana).

Pseudopimelodus; McConnell, 1967, Timehri, 43: 67 (Rupununi).

Material. — Two specimens, before 1864, Essequibo River (Ehrhardt, BM
no. 1864.1.21: 16-17), standard length 117, 119 mm. One specimen, 1908, Kumaka, British
Guiana (C. H. Eigenmann, FM no. 53220), standard length 62 mm, paratype of Pseudo¬
pimelodus villosus. One specimen, 1908, Wismar, British Guiana (C. H. Eigenmann,
FM no. 53571), standard length 39 mm, paratype of Pseudopimelodus villosus. One
specimen, 4 May 1925, Cano de Quiribana into Rio Orinoco, Venezuela (C. Temetz,
SU no. 56987), standard length 84 mm. One specimen, 13 May 1925, Cano de Quiribana
into Rio Orinoco, Venezuela (C. Ternetz, SU no. 56988), standard length 97 mm. One
specimen, May 1925, Cano de Quiribana into Rio Orinoco, Venezuela (C. Ternetz, SU
no. 56968), standard length 81 mm. One specimen, 10 April 1957, Karanambo Creek,
Rupununi (R. H. McConnell log. 3), standard length 107 mm. One specimen, May 1958,
Karanambo Creek, Rupununi (R. H. McConnell), standard length ca. 92 mm. One
specimen, 2 May 1958, Karanambo Creek, Rupununi (R. H. McConnell, BM
no. 1970.10.29: 13), standard length 46 mm. Five specimens, 4 May 1961, Semuni Creek,
Rupununi River, British Guiana (R. H. McConnell, BM no. 1970.10.29: 4-6, 9-10),
standard length 43, 69, 109, 146, 166 mm.

Characters. — Morphologically similar to P. r. raninus, but appears to
attain a larger size and differs strikingly by its much duller and less attrac¬
tive coloration: the whole body is dark brown-grey, more or less mottled with¬
out a nape-band or any other clear markings; the dorsal fin is blackish, with an
ill-defined paler centre; the other fins are mottled, with dark brown-grey
spots on a pale background; no bands on the caudal fin. Especially ventrals
and caudal fin are therefore very different from P. r. raninus .

A small specimen collected by Mrs. McConnell (BM no. 1970.10.29: 13)
differs from the description given above in that it shows markings on the
body as P. r. raninus, but fainter, and in addition an irregular pale band from
behind D to V. Another specimen (BM no. 1970.10.29: 4, 43 mm standard
length) shows very faintly similar markings. Both specimens have the pale,
spotted fins diagnostic of P. r. villosus .

Distribution. — Known from the Demerara River (Kumaka, Wismar),
and the Essequibo basin (Essequibo River, Curupucari [= Kurupukari],

212

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Potaro Landing, Rupununi), British Guiana, and from the Cano de Quiri-
bana, Orinoco, Venezuela. Recently discovered in western Suriname, see
Introduction.

Discussion. — As regards the presumed difference in size, it should be
noted that the largest specimen out of a series of 61 P. r . raninus from Suri¬
name and French Guiana has a standard length of only 82 mm, whereas out of
15 specimens of P. villosus which I examined, six are well over 100 mm in
standard length. Naturally, I have considered the possibility that the differ¬
ence in coloration would be a matter of size (as suggested by Schultz, 1944a:
201, for P. v. butcheri), but this is contradicted by the material, as several
small specimens of P. villosus have been available, all of which have the
ventral fins and caudal fin as described, and unlike P. r. raninus. Also, in the
material of P. r. raninus from Suriname, there is no difference between small
and larger specimens which might be suggestive of a transition of one colour
type into another (large Amazonian specimens have been discussed on a
preceding page).

Eigenmann (1912: pi. X) has mistakenly figured this subspecies with a
rounded caudal fin, although in the text he correctly described the caudal fin
as notched. Also he has described and figured a distinct notch in the caudal
fin of P. albomarginatus, which actually is scarcely or not at all present
(the upper rays of the caudal fin are longer than the lower rays, but there
is no notch). This has clearly contributed to the later confusion between
the two.

Pseudopimelodus raninus acanthochiroides Giintert

Pseudopimelodtts acanthochiroides Giintert, 1942, Zool. Anz., 138: 29 — Santander in
Kolumbien.

Pseudopimelodus villosus butcheri Schultz, 1944, Proc. U.S. Nat. Mus., 94: 199 —
Rio San Juan near bridge south of Mene Grande, tributary to Rio Motatan, Maracaibo
Basin.

Pseudopimelodus villosus butcheri ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool.,
33: 28 (rio San Juan, bacia do Maracaibo, Venezuela) ; Gomes, 1946, Occ. Pap. Michigan
Mus., 494: 7 (Rio San Juan, near bridge south of Mene Grande, Venezuela, Maracaibo
basin).

Pseudopimelodus albomarginatus ; Fernandez & Martin, 1953, La Region de Perija
y sus Habitantes: 306 (Rio Negro, Tokio, a 16 kms oeste de Machiques).

Pseudopimelodus raninus ; Mago-Leccia, 1967, Bol. Soc. Venez. Cienc. Nat., 27: 235
(Venezuela).

Pseudopimelodus villosus villosus ; Mago-Leccia, 1967, Bol Soc. Venez. Cienc. Nat.,
27: 255 (Venezuela).

Material. — One specimen, 1925, Santander, Colombia (E. Ritter, NMB no. 5277),
standard length 107 mm, syntype of Pseudopimelodus acanthochiroides . One specimen,
same data (NMB no. 5278), standard length 67 mm, syntype of Pseudopimelodus acantho¬
chiroides. Three specimens, 17 March 1942, Rio San Juan at bridge south of Mene Grande,

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

213

tributary of Rio Motatan, Maracaibo basin, Venezuela (L. P. Schultz, USNM
no. 121273), standard length 88, 124, 162 mm, paratypes of Pseudopimelodus villosus
butcheri. Three specimens, 20 March 1942, Rio San Pedro, Maracaibo basin, Venezuela
(L. P. Schultz, ZMA no. 102122, ex USNM no. 121271), standard length 31, 45, 58 mm,
paratypes of Pseudopimelodus villosus butcheri. Three specimens, same data (USNM
no. 121271), standard length 35, 37, 46 mm, paratypes of Pseudopimelodus villosus but¬
cheri. Three specimens, 21 March 1942, Rio Machango, 20 km above bridge south of
Laganillas, Maracaibo basin, Venezuela (L. P. Schultz, USNM no. 121276), standard
length 43, 49, 55 mm, paratypes of Pseudopimelodus villosus butcheri.

Diagnostic characters. — Morphologically similar to P. r. raninus, but
body a little more slender, its greatest width 3.35-3.65 times in standard
length. Differs slightly but constantly in colour pattern as follows; there is
a more or less distinct pale band across the body, just behind the dorsal fin
(absent, or occasionally just indicated in P. r. raninus), pectorals with a
variable but considerable amount of white (mostly black in P. r. raninus );
ventrals with a conspicuous white margin (all black in P. r. raninus) ; adipose
fin entirely dark, with a white dot on its origin and sometimes the suggestion
of a white margin (the whole posterior half white in P. r. raninus).

Large specimens lose the colour pattern of the body, but that of the fins
remains distinct and diagnostic at all sizes, as correctly pointed out by
Schultz (1944a: 201). This will also serve to distinguish the present sub¬
species from P. r. villosus. The largest specimen has the adpressed dorsal
fin only reaching 0.6 of the distance to the adipose fin.

Distribution. — The Maracaibo basin in Venezuela and adjacent Colombia
(Santander del Norte) 1 ).

Discussion. — The two syntypes of P. acanthochiroides are much
decolourized, but what remains of their colours and pattern suggests strongly

1) In the description of P. acanthochiroides , Giintert does not make clear if the locality
Santander pertains to the town of that name, Prov. Cauca, southern Colombia, or to the
Province of Santander, but from references on later pages (Giintert, 1942: 38, 39) it
is clear that the Province of Santander is meant. The province of Santander is part of
the Magdalena basin, but a large area in the adjacent Province of Norte de Santander
belongs to the Maracaibo basin, and it appeared likely that the specimens had been col¬
lected there, particularly as there do not appear to be any (other) records of the species
from the comparatively well-known Magdalena basin (cf. Miles, 1947). This has now
been confirmed by Dr. E. Sutter, who informed me as follows (in litt., 2.IV.1971) :
“Herr Dr. H. G. Kugler, Vorsteher der Geologischen Abteilung unseres Museums, teilt
mir mit, dass Herr Dr. E. Ritter um 1925 als Chefgeologe des “Oeldistrictes” Maracaibo
tatig war. Mit “Santander” ist die Provinz gemeint, und zwar zweifellos Santander del
Norte, d.h. die von Maracaibo aus direkt erreichbaren Gebiete Kolumbiens, die zum
Untersuchungsgebiet von Maracaibo gehorten. Nach Dr. Kugler, der selbst viele Jahre
in Siidamerika tatig war, haben Geologen Fische und anderes Material nur an Orten
gesammelt, wo sie ein Standlager hatten und langere Zeit arbeiteten; auf Reisen ware
es zu umstandlich gewesen, Alkoholmaterial zu sammeln. Es ist also ganz unwahrschein-
lich, dass die Fische aus einem andern Teil der Provinz Santander stammen”.

214

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

that they do not differ from fishes described two years later as P. villosus
butcheri. As mentioned in a footnote, the type-locality of P. acanthochiroides,
like that of P. v, butcheri, is the Maracaibo basin.

Pseudopimelodus raninus transmontanus Regan

Pseudopimelodus transmontanus Regan, 1913, Ann. Mag. Nat. Hist., (8) 12: 467 —
the Condoto, the San Juan and the Tamana, and the Durango, W. Ecuador.

Pseudopimelodus transmontanus ; Eigenmann, 1922, Mem. Carnegie Mus., 9: 33, pi. II
fig. 1, 2 (Creeks near San Lorenzo, Rio Telembi, Patia Basin); Gosline, 1941, Stanford
Ichth. Bull., 2: 85 (San Juan and Patia basins, Colombia); Gosline, 1945, Bol. Mus. Nac.
Rio de Janeiro, Z00L, 33: 28 (Colombia e oeste do Equador); Gomes, 1946, Occ. Pap.
Mus. Michigan, 494: 4, 7 (Rio Telembi, near San Lorenzo, Colombia, Patia basin);
Ovchynnyk, 1968, Zool. Anz., 181: 254 (Rio Durango, W. Ecuador).

Material. — One specimen, 14 January 1913, creeks near San Lorenzo, Patia Basin,
Colombia (A. Henn and C. Wilson, CAS no. 17977), standard length 80 mm. One
specimen, 17 January 1913, Telembi above Barbacos, Colombia (A. Henn and C. Wilson,
IUM no. 13008), standard length 57 mm, specimen figured by Eigenmann, 1922. One
specimen, 17 January 1913, creek near San Lorenzo, Rio Telembi, Patia Basin, Colombia
(A. Henn and C. Wilson, FM no. 56518), standard length 100 mm. One specimen, Sep¬
tember 1945, Pizarro, Choco, Pacific side, Colombia (K. von Sneidern, FM no. 70328),
standard length 125 mm.

Characters. — Morphologically identical with P. r . raninus, but colour
pattern (white markings) different and diagnostic. The differences are the
following; the white band across the nape is wider and more even; there
is a conspicuous white spot just before the origin of the dorsal fin (larger
than in P. r. acanthochiroides) ; there is a large irregularly shaped white area
on the flanks, from below the end of the dorsal base diagonally downwards
and backwards; on the origin of the adipose fin is a white dot, as in the
nominate race, but the remainder of the adipose fin is entirely dark, not its
distal half white as in P. r. raninus or with a small white tip as usually in
P. r. acanthochiroides ; the pale band across the caudal peduncle is not well-
developed; the anal fin is dark with a white margin, but it lacks a white median
cross-bar, and the anterior four rays of the ventrals have broad white (hya¬
line) tips (all dark in P . r. raninus).

The largest specimen (FM no. 70328) differs from all others in that the
pale markings on the body are only faintly indicated; the ventral fins have
only their basal halves dark, the whole distal halves hyaline, and all other
fins have also broader hyaline edges then the smaller specimens; the adipose
fin has a small white tip. I presume that the faintness of the body-markings
is a matter of size, and that the difference in fin-pattern is due to individual
variation as it is unlikely that fishes from the Rio Baudo (Pizarro) would
differ from those of the R. San Juan, which is the next important river down
the coast.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

215

Distribution. — Rivers of the Pacific drainage of the Andes in Colombia
(R. Baudo, R. San Juan and R. Patia basins) and Ecuador (R. Durango);
the last-mentioned locality according to Regan (1913); I have failed to find
it on a map.

Pseudopimelodus fowleri (Haseman)

Cephalosilurus fowleri Haseman, 1911, Ann. Carnegie Mus., 7: 317, fig. 1, pi. XLVI —
Cidade da Barra, Bahia, Rio Sao Francisco.

Cephalosilurus fowleri ; Fisher, 1917, Ann. Carnegie Mus., 11: 411 (Penedo); Gosline,
1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 27 (Sao Francisco); Fowler, 1951, Arq.
Zool. S. Paulo, 6: 519 fig. 338 (Rio Sao Francisco).

Lophiosilurus fowleri; A. de Miranda Ribeiro, 1918, Rev. Mus. Paul, 10: 729 (Rio
S. Francisco, Bahia).

Material. — One specimen, 20 March 1908, Penedo (J. D. Haseman, FM no. 58006),
standard length 152 mm.

Diagnostic characters. — Head broad and depressed, somewhat angular at
the gape; pectoral fins with six rays; maxillary barbels short, not quite
reaching to pectoral base; the dorsal fin when depressed covers only three-
fifths of its distance from the adipose fin; caudal fin rounded, not long,
probably pale, mottled with dark spots; occipital process rather narrow,
with a small fork at its end, where it meets the slender tip of the predorsal
plate; gill-rakers moderately developed, fairly numerous but short.

Description. — D 1 . 6 , A 11 (iii.8), P 1 . 6 , V 6 (i.5), C with 15 divided rays;
gill-rakers 1 + 1 + 11, of which only the first three developed, the last ten
quite rudimentary; branchiostegals 9; teeth in both jaws in a broad band, the
band of premaxillary teeth with distinct backward projections laterally; eyes
small, their greatest diameter 3.5 times in snout, 4 times on bony interorbital,
and 13.5 times in head; the maxillary barbels reach about 0.9 of the distance
from their origin to the pectoral base; the outer pair of mental barbels is
about three-fifths of their length, and the inner pair is no more than half
the length of the outer pair.

Head 3.2 times in standard length, greatest width of body 3.35 times in
standard length, greatest depth of body 5.0 times in standard length, pre¬
dorsal length 2.5 times in standard length.

Dorsal fin with one spine and six divided rays, the spine short, equal to
snout and half an eye’s diameter, and 4.8 times in predorsal length, without
serrations; soft rays comparatively short, when pressed down they cover
little more (about 0.55) than half the distance to the adipose fin. Anal fin
with three simple and eight divided rays, its longest rays half the length of
the head. Pectorals with one spine and six rays; the spine is strong, flat,
clearly longer than snout with eye, and 2.9 times in predorsal length; along
the anterior edge is a series of 17 hooks, along its posterior edge a series of

216

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

14 larger ones; the spine has a filament of about half its length; the first ray
equals in length the spine with filament, the second ray is almost as long as
the first, and the last one is a little over half the length of the first ray.
Ventrals rounded, with six rays of which the first one is simple, the others
divided, well-developed, but reaching no more than two-thirds of the distance
to the anal origin, and the longest rays about twice in head. Caudal fin
rounded, not very long, with fifteen divided rays, supported by a number of
simple, more or less rudimentary rays, the longest rays equal to head without
snout, and 4.5 times in standard length. Adipose fin of moderate size, im¬
planted opposite the anal fin; its height about one-third of the depth of the
caudal peduncle, its base 7 times in the standard length.

Colours. The specimen examined is much decolourized, but it looks as if
the head has been pale with small dark spots, whereas most of the body is
pigmented, the fins are pale, distally darker, and the caudal fin has a number
of dark spots. Haseman (1911: 318) described his specimen, collected four
years previously, as: “above and below brown; many black spots on the sides,
back, and fins”.

Distribution. — On present evidence confined to the lower course of the
Rio Sao Francisco, where known from two specimens; the type from Barra,
and the specimen from Penedo examined by me.

Pseudopimelodus albomarginatus Eigenmann (figs. 37c, 38c, pi. 10)

Pseudopimelodus albomarginatus Eigenmann, 1912, Mem. Carnegie Mus., 5: 153, pi. XI
fig. 1 — Tukeit, British Guiana.

Pseudopimelodus albomarginatus Eigenmann, 1910, Rep. Princeton Univ. Exp. Pata¬
gonia, 3: 384 — nomen nudum (Potaro River, British Guiana).

Pseudopimelodus albomarginatus ; Gosline, 1941, Stanford Ichth. Bull., 2: 85 (British
Guiana); Schultz, 1944, Proc. U.S. Nat. Mus., 94: 198, 202 (British Guiana); Gosline,
1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 28 (Guiana Inglesa); Gomes, 1946, Occ.
Pap. Michigan Mus., 494: 6 (Tukeit, Lower Potaro River, British Guiana) ; van der
Stigchel, 1946, South American Nematognathi: 45 (Tukeit, Lower Potaro) ; van der
Stigchel, 1947, Zool. Meded., 27: 45 (Tukeit, Lower Potaro) ; Grey, 1947, Fieldiana,
Zool., 32: 179 (British Guiana: Tukeit, lower Potaro River).

Material. — One specimen, 1908, Waratuk (C. H. Eigenmann, FM no. 53572), standard
length 70 mm. Eight specimens, 1908, Tukeit (C. H. Eigenmann, FM no. 53222), standard
length 37, 39, 46, 60 mm, and four too small and shrivelled for correct measuring. One
specimen, 1908, Tukeit (C. H. Eigenmann, FM no. 7379), standard length 29 mm. One
specimen, 1908, Tukeit (C. H. Eigenmann, ZMA no. 104667), standard length 28 mm.
One specimen, 1908, Tukeit (C. H. Eigenmann, ZMB no. 17913), standard length 35 mm.
One specimen, 1908, Tukeit (C. H. Eigenmann, SU no. 21931), standard length 30 mm.
All these specimens are paratypes. Two specimens, November 1959, Potaro River (R.
Liley, in Mrs. McConnell’s collection, one of these is now RMNH no. 26745), standard
length 83, 85 mm.

Diagnostic characters. — Head very broad and depressed, angular at the

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

217

gape, above rather dark, somewhat mottled; pectoral fins with six rays;
maxillary barbels short, shorter than head and usually not even reaching pec¬
toral base; the dorsal fin when depressed does not quite reach to the adipose
fin; caudal fin long, rounded-lanceolate, black, with a narrow white margin
and a conspicuous irregularly shaped white area near its base; occipital
process and predorsal plate strongly connected; gill-rakers well-developed,
slender.

Description. — D 1 . 6 , A 10-11 (ii or iii.7 or 8), P 1 . 6 , V 6 (i.5), C 11-13
(branched rays only; counted from five specimens only, all others had the
caudal fin damaged or missing), gill-rakers 2 or 3 + 1 + 7, well-developed
and fairly slender; branchiostegals 9 or 10; teeth in both jaws in a broad
band, the band of premaxillary teeth with distinct projections laterally (fig.
37c); eyes very small, their greatest diameter 3-4 times in snout, 3-4.5 times
in bony interorbital, and 11-13.5 times in head; the maxillary barbels usually
do not reach to the pectoral base, the outer pair of mental barbels is about
two-thirds of their length, and the inner pair is one half to two-thirds the
length of the outer pair.

Head 2.5-27 times in standard length, greatest width of body 3.0-3.2 times
in standard length, greatest depth of body 5.0-5.5 (one specimen 6.5) in
standard length, predorsal length 2.2-2.35 times in standard length.

Dorsal fin with one spine and six divided rays, the spine moderately strong
and comparatively short, equal to snout, and 4.1-4.8 times in predorsal length,
without serrations; soft rays long, when pressed down the rays cover about
four-fifths of the distance to the adipose fin. Anal fin with two or three
simple and seven or eight (once nine) divided rays, its longest rays a little
less than half the length of the head. Pectorals with one spine and six rays;
the spine is very strong, flat, a little longer than snout with eye, with on both
sides a series of some 8-14 hooks (the number appears to increase with size:
in the smallest specimen there are no more than 8 on each side, in the largest
specimen 14), those on the posterior margin antrorse, those on the anterior
margin basally antrorse, distally retrorse; the first three rays are about one-
fifth longer than the spine, the fourth equals it in length and the fifth and
sixth rays are much shorter. Ventrals rounded, with six rays, of which the
first one is simple, the others divided, relatively long, reaching to the origin
of the anal fin and the longest rays equal to the distance from the gape to the
pectoral spine. Caudal fin long, lanceolate or lanceolate-rounded, with 11-13
divided rays and several more or less rudimentary simple rays, the longest
rays equal to head without snout, or head without half the length of the
snout. Adipose fin of moderate size, implanted opposite the anal fin; its height
one-third to two-fifths of the depth of the caudal peduncle.

2 l8

ZOOLOGISCHE VERHANDELINGEN I32 (I974)

Colours. A rather complicated pattern of blackish brown on a light brown
background; the largest dark areas are around and below the dorsal fin,
dorsally connected with a large patch below the adipose fin, and on the caudal
base, narrowly connected on each side along the lateral line with the preceding
patch. Upper surface of head slightly mottled, flanks and belly dotted with
brown. Dorsal fin black with a white window across the last five rays and
a narrow white margin; anal fin with a white base and white margin; pectoral
fins with a white base and small white tips; ventrals black with a white base
and a broad white margin; adipose fin light brown, with in the middle an
area of blackish brown; caudal fin blackish, with an irregular white area near
its base, largest on the median rays, and a comparatively broad white margin.

Distribution. — Hitherto only known from the Potaro River, a tributary
of the Essequibo. It is surprising, and may be significant, that the species
is not known from the Rupununi, another tributary of the Essequibo, which
is at least as well-known ichthyologically as the Potaro.

Discussion. — There has been doubt about the validity of P. albomargi-
natus ; Schultz (1944a : 198) thought that it was probably the same as P. villo -
sus\ he also stated that the shape of the caudal fin, in which these two species
differ, is not a reliable character as it changes with growth. Fernandez &
Martin (1953) went even farther and decided that P, villosus butcheri (= P.
raninus acanthochiroides of this revision) is a synonym of P. albomarginatus ;
they made no mention of P. villosus.

The examination of several paratypes, and of two more recent specimens in
an excellent condition of preservation, has enabled me to confirm that P.
albomarginatus is a valid species, probably more closely related to P. nigri-
cauda and P. fowleri than to P. raninus.

Pseudopimelodus nigricauda species nova (figs. 37c!, 38d, pis. 11, 12)

Material. — Thirteen specimens as follows. One specimen, 23 September 1965, Zand-
vallen, Kabalebo (Mees, RMNH no. 26736), standard length 29 mm. One specimen, 27
November 1965, Tapanahoni, about 2 km upstream of its confluence with the Paloemeu
(Mees, RMNH no. 26737), standard length 49 mm. Two specimens, same date and
locality (Mees, RMNH no. 26738), standard length 171 and 189 mm. One specimen,
23 January 1966, Sipaliwini (Mees, RMNH no. 26739), standard length 212 mm, total
length 271 mm, holotype of the species. One specimen, same data (Mees, RMNH
no. 26740), standard length 61 mm. One specimen, 4 February 1966, Sipaliwini (Mees,
RMNH no. 26741), standard length 77 mm. One specimen, 6 February 1966, Sipaliwini
(Mees, RMNH no. 26742), standard length 65 mm. Two specimens, 9 February 1966,
Sipaliwini (Mees RMNH no, 26743), standard length 70 and 106 mm. One specimen,
30 January 1967, Gran Mau, a small tributary of the Gran Rio about 13 km south of
Djoemoe (H. Nijssen, ZMA no. 105948), standard length 128 mm. One specimen, 21
March 1967, Suriname Rivier, ca. j l / 2 km north of Botopassie (H. Nijssen, ZMA
no. 105949), standard length 143 mm. One specimen, 21 April 1967, Maka Kreek, a

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

219

tributary of the Lawa Rivier, about 10 km south of Stoelmanseiland (H. Nijssen,
ZMA no. 105951), standard length 104 mm.

Diagnostic characters. — Head very broad and depressed, angular at the
gape, above light brownish with a complicated pattern of black markings;
pectoral fins with six rays; maxillary barbels reaching to the base of the
pectoral fins or four-fifths of that distance; the dorsal fin when depressed
reaches to the adipose fin; caudal fin long, lanceolate, black with a narrow
white margin; occipital process and dorsal plate narrowly separated, the
occipital process forked at its end, where it is met by the slender tip of the
predorsal plate; gill-rakers moderately developed, fairly numerous, but short,
almost rudimentary.

Description. — D 1.6 or I. 7, A 9-11 (ii to iv.7 or 8), P 1 . 6 , V i.5, C with
14 to 16 divided rays; gill-rakers 1 or 2+1 + 9 to 13, usually no more
than three on the hypobranchial developed, the others more or less rudi¬
mentary ; branchiostegels 9 or 10; teeth in both jaws in a broad band, the band
of premaxillary teeth with distinct backward projections laterally (fig. 37d);
eyes small, their greatest diameter 2-2.5 times in snout, 3-4 times in bony
interorbital, and 9-10 times in head; the maxillary barbels reach from 4/5
to 1 the distance from their implantation to the pectoral base; the outer pair
of mental barbels is about three-quarters of their length, and the inner pair
is one half to two-thirds the length of the outer pair.

Head 2.8-3.0 times in standard length, greatest width of body 2.7-3-3 times
in standard length, greatest depth of body 4.2-54 times in standard length,
predorsal length 2.3-2.5 times in standard length.

Dorsal fin with one spine and six divided rays; one specimen (no. 26742)
has it with one spine and seven rays; dorsal spine moderately strong and not
very long, about equal to snout with eye, without serrations; soft rays long,
when pressed down the rays reach, or nearly reach, to the adipose fin. Anal
fin with two to four simple and seven or eight divided rays, rounded, its
longest rays about half the length of the head. Pectorals with one spine and
six rays; the spine is very strong, flat, varying from not much over the length
of snout with eye to nearly half the length of the head, and 2.5-2.6 times in
predorsal length, with on both sides a series of about a dozen strong antrorse
hooks, their length about equal to the breadth of the spine; the first ray longer
than the spine, the following rays successively shorter, and the last ray about
two-thirds the length of the first ray. Ventrals rounded, with six rays, of
which the first one is simple, the others divided, relatively long, reaching
almost to the anal fin, and the longest rays equal to the distance from the gape
to the pectoral fin. Caudal fin long, lanceolate or lanceolate-rounded, with
14 to 16 divided rays besides several developed simple rays and many rudi-

220

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

ments, the longest rays equal to head without snout or to head without half
the length of the snout. Adipose fin moderate, implanted opposite the anal fin,
and its base of the same length as that of the anal fin or backwards a fraction
longer, its height about half the depth of the caudal peduncle below it.

Colour. The colours are light brown and black. Generally speaking, the
posterior part of the body, behind a vertical through the dorsal spine, is black,
the anterior part light brownish. The upper surface of the head shows, on the
light brownish background, a complicated though fairly regular pattern of
black, and farther backwards, connecting the upper edges of the gillslits is a
blackish stripe, insufficiently solid to be called a bar, across the nape. On the
(mainly blackish) posterior part of the body, there is a large light brown area
on the flanks between D and adipose fin, which downwards extends forwards
to reach the implantation of V. The caudal peduncle has a light brown or
whitish saddle, that also takes in the hind border of the adipose fin. There
is a pale spot on the anterior border of the adipose fin, and another one on
the back just behind the last dorsal ray, which extends as a small triangular
area on the dorsal fin, where it includes a section of the last two rays and
their connecting membrane. The pectoral fins are spotty blackish brown
distally, hyaline with smaller and indistinct greyish brown spots basally; the
dorsal fin is black, except for the pale triangle just mentioned and for very
small white tips to the rays; the anal fin is dark, with some faint lighter areas
and a small but distinct pale spot on the last ray near its base; the ventrals
are dark, distally almost black, but with many paler spots, and the middle
part of the last ray whitish, this white sometimes extending as a wedge to
the one of two preceding rays; the tail behind the saddle is black, with
usually some very narrow and inconspicuous white tips to the longest rays;
this entirely black caudal fin is an excellent character.

Habitat. — All my specimens were obtained in rivers, where they would
be at least periodically exposed to a very strong current, but in the rivers they
stayed in the larger pools rather than in the rapids; their shape suggests
bottom-dwellers.

Distribution. — Widely distributed in Suriname, where known from the
basins of the Marowijne (Tapanahoni, Maka Kreek), the Suriname Rivier,
and the Corantijn (Sipaliwini, Kabalebo). This distribution makes it prac¬
tically certain that the species occurs also in French Guiana, and probably in
British Guiana, but presumably not as far west as the Essequibo (see Dis¬
cussion) .

Discussion. — Presumably P. nigricauda is related to P. albomarginatus ;
the two species agree in general morphology, including body-shape and
development of the occipital process. On the other hand, the differences in

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

221

colour-pattern are considerable, and as in certain morphological characters
(development of the gill-rakers) they differ also, I am of the opinion that
P. nigricauda is a separate species, not a subspecies of P. albomarginatus.
The two species may replace each other geographically; at least, I consider
it unlikely that P. nigricauda would have been overlooked in the comparatively
well-known Essequibo basin, if it did occur there.

As this species has such a wide and strongly depressed head, I have con¬
sidered the possibility that it would be referable to Lophiosilurus alexandri
Steindachner (1876b: 154, pi. XV), but judging by its original description
and figure, the type of L. alexandri differs by its much shorter dorsal fin
which when pressed down covers only a fraction of its distance from the
adipose fin, the very small and low adipose fin, the relatively shorter and
more rounded caudal fin which moreover has only 12 instead of 14-16 divided
rays, and the different pectoral spine, with more teeth. As in the available
material of P. nigricauda there does not appear to be much change of the
listed characters with growth, and the largest specimen is not much smaller
than the type of L. alexandri, the differences noted must be specific. Whether
they are also sufficient to uphold the monotypic genus Lophiosilurus is a
different question, which I am inclined to answer in the negative.

In general appearance P. nigricauda is also dose to P. fowleri, the only
species with which it agrees in shape and number of gill-rakers. As indicated
in the key and in the description, its head is a little broader and larger, its fins
are distinctly longer, and the colour-pattern differs, especially that of the
caudal fin.

Genus Microglanis Eigenmann

Microglanis Eigenmann, 1912, Mem. Carnegie Mus., 5: 155 — type by original designa¬
tion, Microglanis poecilus Eigenmann.

Diagnosis. — Very similar to Pseudopimelodus but can be distinguished by
three characters: small size, shape of the band of teeth in the upper jaw, and
reduction in the posterior part of the lateral line.

In appearance this is a minute edition of Pseudo pimelodus : the same heavy,
somewhat flattened head, the same tapering body; spines and fin-rays
generally very similar in structure and in numbers; barbels reaching to
beyond pectoral base, but not (or only just) reaching to a vertical through
the dorsal origin. When comparing very small specimens of Pseudo pimelodus
with Microglanis poecilus, one would certainly not recognise the latter as
generically different, but the fact remains that Microglanis never grows to a
larger size, the maximum size found in a large series from Suriname (2
species) being a standard length of 38 mm only. The largest species of the

222

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

genus as here defined is M. parahybae, of which I have examined a specimen
of 79 mm standard length.

The premaxillary band of teeth is laterally rounded, without backward
projections, as occur more or less distinctly in all species of Pseudo pimelodus.

The lateral line, although complete, has no distinct pores on the posterior
part of the body, whereas in Pseudo pimelodus it is developed and pore-bearing
over its whole length.

Distribution. — Tropical South America from Venezuela through the
Guianas and the Amazon basin to the Rio de la Plata, and west of the Andes
in Ecuador.

Discussion. — Schultz (1944a) doubted the validity of the genus Micro -
glanis, as he found variation in the main character (shape of patches of
teeth) it was based on, and Myers (in Gomes, 1946: 8) has suggested that
the species described in Micro glanis are only young of Pseudo pimelodus. Hoe-
deman (1961: 135) referred to Schultz with approval as follows: “Further¬
more I must agree with Schultz (1944: 197) where he doubts the validity
of the genus Microglanis , at least of the species Microglanis poecilus . An
18 mm st.l. specimen obtained in exchange from the Carnegie Museum,
entered in our collection under the manuscriptname Batrachoglanis raninus,
and which belongs to the typical series upon which Microglanis poecilus was
established, doubtlessly is a juvenile specimen of raninus tf . In defense of
Schultz it must be said that he was only concerned with the generic name
Microglanis and never expressed doubt about the validity of the species
Microglanis poecilus ; Hoedeman must have misunderstood him. Eigenmann
(1912: 155, footnote) has mentioned that a number of duplicates of Micro¬
glanis poecilus were labelled as Batrachoglanis raninus. This was a mistaken
original identification, of which Hoedeman appears to have been unaware.
Actually Schultz, although doubting the validity of the character, did observe
a difference in dentition as in a series of Pseudo pimelodus villosus butcheri
(= P. raninus acanthochiroides in this revision) he found: “The premaxillary
band of teeth in this series is angular in the small ones 40 to 50 mm in length,
and in the larger ones this angle projects more posteriorly as described for
Pseudo pimelodus. In a small paratype of Microglanis poecilus the outer or
lateral ends of the premaxillary band of teeth are more rounded

I can confirm the characters originally given by Eigenmann; there cer¬
tainly is a difference in dentition between Microglanis and the species of
Pseudo pimelodus as well as between these species inter se (compare fig. 37
and 40). In this connection it is worth recording that, contrary to what
Schultz found in Venezuela, there does not appear to be any difference in
shape of the tooth-patches between large and small specimens of the Suri-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

223

name species. Even where the difference in size is great, for example between
the largest and smallest specimens of Pseudopimelodus nigricauda, the tooth-
patches are of identical shape. Gomes (1946: 8) found the same in several
other species, including Pseudo pimelodus villosus butcheri.

The genus Microglanis as defined above contains seven species: M. parahy -
bae, M . poecilus , M. variegatus, M. ater, M. zonatus, M. iheringi, and M,
secundus. Two species occur in Suriname.

e .—

9

Fig. 40. Right-hand pectoral-spines and dentition of Microglanis. (a, b) M. parahybae,
large specimen, spine 4 X (all other spines 8 X), (c, d) M. poecilus , (e) M. secundus,
(f) M. iheringi, (g) M. variegatus.

Microglanis parahybae (Steindachner) (fig. 40a, b)

Pseudopimelodus Parahybae Steindachner, 1880, Denkschr. Akad. Wiss. Wien,
Mathem.-Naturw. CL, 42 (1) : 60 — aus dem Rio Parahyba und von Santa Cruz.

Pimelodus ( Pseudopimelodus ) cottoides Boulenger, 1891, Proc. Zool. Soc. Lond.: 233,
pi. XXV fig. 2 — Camaquam River, Rio Grande do Sul.

Pseudopimelodus charus ; Steindachner, 1876, Sitzb. Akad. Wiss. Wien, Mathem.-
Naturw. CL, 74 (1): 632 (Rio Parahyba und S. Cruz).

Pseudopimelodus parahybae ; Eigenmann & Eigenmann, 1888, Proc. Calif. Acad. Sci.,
(2) 1: 122 (Santa Cruz); Eigenmann & Eigenmann, 1890, Occ. Pap. Calif. Acad. Sci., 1:
no (Santa Cruz); Eigenmann & Eigenmann, 1891, Proc. U.S. Nat. Mus., 14: 27 (Rio

224

ZOOLOGISCHE VERHANDELINGEN I32 (I974)

Parahyba to Rio Doce); A. de Miranda Ribeiro, 1911, Arch. Mus. Nac. Rio de Janeiro,
16: 253, fig. 104 (Rio Parahyba e affluentes — Santa Cruz); Fisher, 1917, Ann. Car¬
negie Mus., 11: 410 (Rio Ribeira; Sao Joao do Barra, Rio Parahyba; Campos, Rio
Parahyba; Uruguayana).

Batrachoglanis parahybae ; Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3:

383 (Rio Parahyba to Rio Doce).

Pseudo pimelodus cottoides ; Eigenmann, 1910, Rep. Princeton Univ. Exp. Patagonia, 3 :

384 (Camaquam River, Paraguay River); Bertoni, 1914, Fauna Paraguaya: 7 (Para¬
guay); Pearson, 1937, Proc. Calif. Acad. Sci., (4) 23: in (Paraguay basin).

Microglanis parahybae ; Arnold & Ahl, 1936, Fremdl. Siisswasserfische: 248-249, fig.
(das sudostliche Brasilien und das dahinterliegende Paraguay und Argentinien); Gosline,
1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 28 (rio Paraiba); Sterba, 1959, Siiss-
wasserfische aus aller Welt: 302, fig. 624 (Rio Parahyba, Rio Doce); Emmens &
Axelrod, 1968, Catfish: 30 (Brazil); Vogt, 1970, Grzimeks Tierleben, 4: 403 (das siid-
ostliche Brasilien bis nach Argentinien).

Microglanis cottoides ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 28 (Rio
Grande do Sul); Gomes, 1946, Occ. Pap. Mus. Michigan, 494: 15 (Rio Grande do Sul
(probably Rio Camaqua) ; Santa Cruz, Rio de Janeiro; Lagoa dos Quadros basin, Con-
ceiqao do Arroio County, Rio Grande do Sul; Rio Paraiba, Rio de Janeiro; Rio Ribeira,
Sao Paulo; Uruguaiana; Rio Grande do Sul); Fowler, 1951, Arq. Zool. S. Paulo, 6:
530 (Rio Doce, Rio de Janeiro, Sao Paulo, Rio Grande do Sul, Paraguay); Ringuelet,
de Aramburu & Aramburu, 1964, Physis, 24: 369 (Francesca Cue, Formosa, Argentina);
Ringuelet, Aramburu & de Aramburu, 1967, Los Peces Argentinos de Agua Dulce: 302
(Francesca Cue, Fsa.); Lundberg & Baskin, 1969, Amer. Mus. Novit., 2398: 6, 28 (no
locality); Castello, 1971, Physis, 30: 543, 544 (Rio de la Plata).

Microglanis poecilus ; (pt.) Gomes, 1946, Occ. Pap. Mus. Michigan, 494: 15 (?Amazon
basin).

Material. — One specimen, ca. 1874, Santa Cruz near Rio de Janeiro (MV no. 44436),
standard length 50 mm, syntype of Pseudopimelodus Parahybae. One specimen, ca. 1889,
Camaquam R., Rio Grande do Sul (H. v. Ihering, BM no. 1889-8-24: 5), standard length
64 mm, syntype of Pimelodus (Pseudopimelodus) cottoides. Three specimens, ca. 1890,
Rio Grande do Sul (H. v. Ihering, BM no. 1891-3-16: 36-45), standard length 26^2, 28
79 mm, syntypes of Pimelodus (Pseudopimelodus) cottoides. Six specimens, 15 June
1908, Campos, Rio de Janeiro (J. D. Haseman, FM no. 58142), standard length 22-23 mm.
Five specimens, 24 June 1908, Sao Joao da Barra, Rio de Janeiro (J. D. Haseman, FM
no. 58141), standard length 22-24 mm. One specimen, July 1941, vicinity of Rio de
Janeiro (T. D. White, SU no. 36949), standard length 17 mm. One specimen, undated,
San Bernardino, Lagune Ipacaray, Paraguay (Fiebrig, ZMB no. 20037), standard length
37^2 mm. Two specimens, undated, “Amazon (?)” (no origin, AMNH no. 14663),
standard length 66, 67 mm.

Aberrant specimens. One specimen, received 1903, Pedre a molhar (ex. Steindachner,
MV no. 44429), standard length 37 mm. One specimen, received 1908, Joinville (MV
no. 44434), standard length 50 mm. One specimen, received 1910, Pedre a molhar (ex
Steindachner, MV no. 44431), standard length 44 mm.

Characters. — D 1 . 6 , A 12, once 13 (usually iii.9), P I.5 or 1.6, V 6 (i-5),
C 12-15 (branched rays), branchiostegals 9, gill-rakers 6-11 (1 to 3 + 1 + 4
to 7). Postoccipital process well-developed, ending in a V-shape, into which
the anterior tip of the longer predorsal plate is fitted.

Proportions of six large specimens (3 7 J 4 , 50, 64, 66, 67, 79 mm in stan-

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

225

dard length): head 3.2-3.8, greatest width of body 3.0-3.4, greatest depth of
body 3.9-44, predorsal length 2.$-2.g times in standard length. The small
specimens (17-28 mm in standard length) are slightly more slender; head
3-3"3-5> greatest width of body 3.5-4.0, greatest depth of body 4.5-54, pre¬
dorsal length 2.6-3.0 times in standard length. The specimens listed as aber¬
rant are puzzling; they agree with the others in counts and in colour pattern,
but have distinctly smaller heads and are far more slender; head 3.5-3.8,
greatest width of body, 3.8-4.0, greatest depth of body 4.6-54, predorsal
length 2.8-3-2 times in standard length. Note the considerable difference in
body-width between these specimens and the three specimens listed first;
admittedly the small specimens are in certain proportions intermediate, but
that does nothing to explain the differences between the larger individuals.
I have considered the possibility that the slender fishes belong to a different
species, but this would make the complete agreement in counts and colour-
pattern difficult to understand; the difference, therefore, remains unex¬
plained, and requires further investigation.

Small specimens of this species are extremely similar to M. poecilus, but
the number of anal rays is higher (12 against 10), and there are differences
in colour-pattern, the most conspicuous of which is that the dark patch
covering much of the adipose fin is much larger, and extends forwards to, or
nearly to the dorsal fin, and often also backwards to the dark caudal base. In
body-shape and structure of the pectoral spine there is little difference be¬
tween the two species, in individuals of similar size. M. parahybae, of course,
reaches a much larger size than M. poecilus, and these large individuals also
develop a different pectoral spine with more teeth (fig. 40a). Gomes (1946)
distinguished the species by width of the gape: mouth large, the gape (greatest
width across opening mouth) less than 1.6 in head length, M. poecilus, and
mouth small, the gape more than 1.7 in head length, M. cottoides (= M. para¬
hybae). However, two syntypes (of M. parahybae and M. cottoides) have
the gape 1.5 and 1.5 to 1.6 in the head, so that this character is unreliable.
Probably Gomes studied slender “aberrant” individuals only.

Distribution. — South-eastern Brazil: rivers running east from the Rio
Parahyba southwards, and Uruguaiana, Rio Uruguai. Argentina: a single
specimen known, from Francesca Cue, Rio Pilcamayo basin, Formosa 1 ).

1) Recently this species was recorded as “relativemente abundante” near Buenos Ayres
(Castello, 1971). Dr. Castello has been so kind to present me with two specimens which
he collected personally at the Obras Sanitarias de la Nation, Buenos Ayres, 16.VI.1965
and 8.IX.1965, standard length 45, 54^ mm (RMNH no. 26705). These specimens agree
in proportions and counts with the six large specimens described in the text, except that
they have An (iii.8), and the numbers of gill-rakers are 3 + 1+8 and 3 + 1 + 10
(the last two rudimentary).

15

226

ZOOLOGISCHE VERHANDELINGEN 13 2 (1974)

Paraguay: Lagune Tpacaray. The species is almost certain to occur in Uru¬
guay, 'but all records hitherto published are evidently based on misidentified
specimens of Pseudopimelodus zungaro mangurus. From the Eigenmanns
(1890) onwards, the species has also been listed as occurring in the Rio Doce,
but I believe mistakenly, for nobody has ever mentioned material from that
river.

Discussion. — Gomes (1946: 15) included in M. poecilus two specimens
of uncertain provenance (? Amazon), with standard lengths of 68 and 69 mm
(AMNH no. 14663). As this is almost twice the maximum standard length
found by me in a material of well over 700 specimens of that species, I
borrowed the specimens to check their identification. Whereas what remains
of their colour pattern is not unlike M. poecilus, they can be distinguished at
once by their entirely different pectoral spines. This spine has its bend not
at about one-third of its length from base, but at about two-thirds. More¬
over, instead of less than 15, there are about 30 teeth along its anterior edge,
and all these teeth point outwards in the same direction, not some outwards
and others inwards as in M. poecilus. Along the posterior edge there are
about 22 larger, antrorse hooks. In all their characters these specimens agree
with M. parahybae of similar size, to which species they are undoubtedly
referable. The suggested provenance “? Amazon” is probably wrong and
should be south-eastern Brazil.

Gomes (1946) used for this species the name Pimelodus {Pseudopimelo¬
dus) cottoides Boulenger, 1891, instead of Pseudo pimelodus parahybae Stein-
dachner, 1880, although the latter has priority. The legal base of the rejection
of the last-mentioned name would be found in article 59 (b) of the present
Code: “A species-group name that is a junior secondary homonym must be
rejected by any zoologist who believes that the two species-group taxa in
question are congeneric”.

The facts are as follows: there are a Pimelodus ( Rhamdia ) parahybae
Steindachner, 1876, and a Pseudopimelodus parahybae Steindachner, 1880.
Boulenger (1891) described Pimelodus {Pseudopimelodus) cottoides, and
added: “The nearest ally of P. cottoides would then be Pseudopimelodus
parahibae, Stdr. [lapsus for parahybae !], with which it is possibly identical.
The proposal of a new name is, however, justified, even should the two species
be the same, as the name parahibae is preoccupied in the genus Pimelodus for
a species of the subgenus Rhamdia {R. parahibae, Stdr.)”.

Nobody before or since Boulenger has ever regarded Pimelodus {Rhamdia)
parahybae Steindachner and Pseudopimelodus parahybae Steindachner as
congeneric. Moreover, Pseudopimelodus parahybae and Pimelodus {Pseudo¬
pimelodus) cottoides have been treated as different species, both names having

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

227

been used and regarded as valid at least until 1945, when Gosline (1945: 28)
listed them as Microglanis parahybae and Microglanis cottoides.

The synonymy of the two nominal species was accepted by Gomes (1946:
16), who gave the following comment:

“As a consequence of Boulenger’s action (1891: 233) in uniting Stein-
dachner’s species, Pimelodus ( Rhamdia) parahybae (1876) and Pseudopime-
lodus parahybae (1880) in the genus Pimelodus, the latter specific name
becomes a homonym, and is not available even though the species was sub¬
sequently placed in Microglanis. Boulenger correctly indicated that the name
cottoides was necessary even if the nominal species parahybae and cottoides
should prove to be identical. It is now apparent, after the study of the speci¬
mens indicated below, that Pseudo pimelodus parahybae Steindachner, 1880,
and Pimelodus {Pseudopimelodus) cottoides Boulenger, 1891, are the same,
and the species must be known as Microglanis cottoides (Boulenger)

In my opinion, objections can be made against this conclusion, the most
important one of which is that Boulenger did not propose P. cottoides as a
nomen novum for P. parahybae, but only suggested that they might possibly
be the same, and that in the following 55 years both names have remained in
use, and have been considered to apply to different species. It was therefore
not Boulenger (1891) but Gomes (1946) who definitely rejected P. parahybae
as a homonym. But in doing so Gomes was not covered by article 59, as he
did not regard Pimelodus {Rhamdia) parahybae and Pseudo pimelodus para¬
hybae as congeneric.

The case presented above clearly shows how article 59 (b) can create
problems, and, as I have previously stated (Mees, 1957, stelling VI), my
own opinion is that nomenclatural stability and consistency would be better
served by the rule that a name, rejected as a secondary homonym, is to be
restored if the condition of secondary homonymy ceases to exist (or is to be
retained by those who did not accept its existence in the first place), even if
the original rejection took place before i960.

I agree with Gomes that P . parahybae and P. cottoides are synonymous;
the preceding discussion was necessary to explain why I regard P. parahybae
as a valid name, and see no need to replace it.

Specimen MV no. 44431 yielded an isopod, which was attached to the
ventral surface just behind the anus. The specimen was examined by Dr.
L. B. Holthuis, according to whom it is a juvenile, belonging to the Cymo-
thoidea.

228

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Microglanis poecilus Eigenmann (figs. 40c, d, 41, pi. 13)

Migroglanis poecilus Eigenmann, 1912, Mem. Carnegie Mus., 5: 155, pi. XII fig. 2 —
below Packeoo Falls, British Guiana.

Pseudopimelodus villosus ; (pt.) Fisher, 1917, Ann. Carnegie Mus., 11: 140 (Santarem:
the smallest specimen only).

Microglanis poecilus; Caporiacco, 1935, Monit. Zool. Ital., 46: 58 (Campo i° Deme-
rara) ; Gosline, 1941, Stanford Ichth. Bull., 2: 85 (British Guiana); Eigenmann & Allen,
1942, Fish. W. South America: 89 (no locality) ; Schultz, 1944, Proc. U.S. Nat. Mus.,
94: 197 (no locality); Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 28
(Guiana Inglesa); (pt.) Gomes, 1946, Occ. Pap. Mus. Michigan, 494: 15 (Packeoo Falls
in Essequibo River; Cano de Quiribana, Venezuela); Fowler, 1951, Arq. Zool. S. Paulo,
6: 531 (Amazonas ?, Guiana, Venezuela); Boeseman, 1953, Zool. Meded., 32: 2, 3
(Djaikreek); Axelrod & Schultz, 1955, Handb. Trop. Aquarium Fish.: 347, fig. (British
Guiana; Rio Apure and Rio Orinoco, and Amazon basin); Lowe, 1964 J. Linn. Soc.,
Zool., 45: 116, 140 (Rupununi); Mago-Leccia, 1967, Soc. Venez. Cienc. Nat. Bol., 27:
255 (Venezuela); Emmens & Axelrod, 1968, Catfish: 28 fig. (Guyana).

Material. — 714 specimens from Suriname as follows. One specimen, 8 October 1948,
Djaikreek (D. C. Geijskes & P. Creutzberg, RMNH no. 19457), standard length 23 mm.
45 specimens, same data (RMNH no. 23876), standard length 16^4-25 mm. 59 specimens,

17- 18 February 1964 Jabocai and Witte Kreek, creeks between Kabel and Lombe (Boe¬
seman, RMNH no. 26073), standard length 20-31 mm. Two specimens, 28 May 1964
shore of Brokopondo Meer, S.E. of Affobakka (Boeseman, RMNH no. 26074), standard
length 28, 29 mm. Ten specimens, 30 July 1964, tributary of middle course of Gran
Kreek (Boeseman, RMNH no. 26075), standard length 24-31 mm. Nine specimens,
7 April 1965, Para Kreek (Mees, RMNH no. 26076), standard length 24-28 mm.
16 specimens, 20 April 1965, Para Kreek (Mees, RMNH no. 26077), standard length

18- 30 mm. Eight specimens, Para Kreek, 27 April 1965 (Mees, RMNH no. 26078),
standard length 26-29 mm - Two specimens, 22 June 1965, Gansee, Brokopondo Meer
near its shore (Mees, RMNH no. 26079), standard length 29, 30^2 mm. Two specimens,
16 July 1965, creek near Kajana, Gran Rio (Mees, RMNH no. 26080), standard length
27, 29 mm. 101 specimens, 28 December 1965, Marshall Kreek (Mees, RMNH no. 26081),
standard length 22-32 mm. 50 specimens, 28 December 1965, Marshall Kreek (Mees,
RMNH no. 26082), standard length 20-30 mm. Five specimens, 28 December 1965,
Marshall Kreek (Mees, RMNH no. 26083), standard length 23-30 mm. 118 specimens,
28 December 1965, Marshall Kreek (Mees, RMNH no. 26084), standard length 21-
32 mm. One specimen, 12 March 1966, tributary of the Mamakreek near Berg-en-Dal
(Mees, RMNH no. 26085), standard length 30 mm. Four specimens, 6 May 1966, Gran
Kreek 50^2 km S. of Affobakka (Nijssen, ZMA no. 106009), standard length 24, 24, 2554,
27 mm. Twelve specimens, 27 May 1966, Gran Kreek, 51 km S. of Affobakka (Nijssen,
ZMA no. 106011), standard length 23^-31, 37 (!) mm. One specimen, 8 June 1966,
Gran Kreek, 60 km S. of Affobakka (Nijssen, ZMA no. 106010), standard length 28 mm.
Two specimens, 8 June 1966, Gran Kreek, 41 km S. of Affobakka (W. Vervoort, RMNH
26527), standard length 26^4, 30 mm. One specimen, 9 June 1966, Gran Kreek, 58 km S.
of Affobakka (Nijssen, ZMA no. 106003), standard length 30 mm. 20 specimens, 8 De¬
cember 1966, Marshall Kreek (Nijssen, ZMA no. 106368), standard length 21-33 mm.
16 specimens, 14 December 1966, Carolina Kreek (Nijssen, ZMA no. 105551), standard
length 22-26 mm. Three specimens, 28 January 1967, tributary of the Gran Rio, 2}4 km
N. of Awarradam (Nijssen, ZMA no. 106008), standard length 28, 28 28 mm. Two
specimens, 1 February 1967, tributary of Gran Rio, 3 km N. of Awarradam (Nijssen,
ZMA no. 105544), standard length 29, 29 mm. 23 specimens, 27 February 1967, tributary
of Kleine Saramacca, 11 km from the mouth of the latter (Nijssen, ZMA no. 105665),

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

229

standard length 24-28 mm. 26 specimens, 28 February 1967, tributary of Kleine Sara-
macca, 13 km from the mouth of the latter (Nijssen, ZMA no. 105646), standard length
22-28^2 mm. Two specimens, 18 March 1967, Awara Kreek near Botopassie (Nijssen,
ZMA no. 106370), standard length 27, 28^4 mm. 31 specimens, 19 March 1967, Kwati
Watra Kreek near Botopassie (Nijssen. ZMA no. 106005), standard length 24-32 mm.
25 specimens, 20 March 1967, Parwapa Kreek near Botopassie (Nijssen, ZMA no. 105790),
standard length 23-31 mm. Nine specimens, 21 March 1967, Jenfee Kreek, 7^4 km N.
of Botopassie (Nijssen, ZMA no. 105695), standard length 25-29 mm. 62 specimens,
15 April 1967, tributary of the Nickerie Rivier, 12 km S. of Stondansie (Nijssen, ZMA
no. 105830), standard length 20^-3154 mm. Five specimens, 17 May 1967, headwaters of
Coppename Rivier, 3°48'N., 56°57'W. (Nijssen, ZMA no. 106006), standard length 27,
2734, 29, 30, 3154 mm. 32 specimens. 18 May 1967, tributary of Coppename, 3°52'N.,
56°55 # W. (Nijssen, ZMA no. 106007), standard length 24-32 mm. Nine specimens,
10 August 1968, creek near aerodrome, Kaysergebergte (M. S. Hoogmoed, RMNH
no. 26528), standard length 17-23 mm.

Extra-limital material. One specimen, 15-19 December 1909, Santarem, Brazil (J. D.
Haseman, FM no. 58140), standard length 22^4 mm. Six specimens, May 1925, Cano de
Quiribana into Rio Orinoco, some 24 km north of mouth of Rio Apure, Venezuela (C.
Temetz, SU no. 40190), standard length 22^-28 mm. One specimen, same data (SU
no. 58577), standard length 21 mm. 15 specimens, same locality, date and collector not
given but presumably the same as above (SU no. 40189), standard length 20-25 mm. Two
specimens, 17 September 1957, Morebay pond Creek, Rupununi, British Guiana (R. H.
McConnell, BM no. 1971.7.29: 31-32), standard length 15, 16 mm. One specimen, 24 Sep¬
tember 1957, Karanambo Creek, Rupununi (R. H. McConnell, BM no. 9171.7.29: 33),
standard length 18 mm. 16 specimens, 4 May 1961, Semuni tacubas, Rupununi (R. H.
McConnell, BM no. 1971.7.29: 34-49), standard length 17-24 mm.

Characters. — D 1.5 or 1 . 6 , A 10 (usually iii.7), P 1.5, V 6 (i.5), C 10-13
(divided rays), and several shorter and rudimentary simple rays, branchio-
stegals 9, gill-rakers 1 or 2 + 1 + 3 to 5. Pectoral spine (fig. 4od) with a
sharp point and usually an Y-shaped tooth along its anterior edge, as described
under the next species; caudal fin somewhat variable, usually scalloped with
rounded (or at least not very sharp) lobes, the upper one the larger.

Colours. Dark earth-brown markings on a light brown background. The
colour pattern, notwithstanding some slight individual variation, is extremely
characteristic, as indicated on the plate. The pale band across the nape is
irregular, running zig-zag, and in a few specimens it is actually interrupted,
the dark areas on head and predorsal region being in contact across it. The
predorsal-dorsal dark area is of a peculiar four-cornered shape as shown
excellently by Eigenmann (1912; pi. XII fig. 2). The dark band across the
caudal base is always more or less triangular in shape, with its apex directed
forwards; a rudimentary dark band from the apex to the posterior part of
the adipose fin is sometimes present. The dorsal as well the caudal fin have
a dark cross-bar. The dark markings of the body are enhanced by the fact
that just around them the background is paler than elsewhere on the body,
almost white instead of light brown.

The coloured figure published by Axelrod & Schultz (1955) and Emmens

230

ZOOLOGISCHE VERHANDELINGEN 132 (I974)

Fig. 41. The distribution of Microglanis poecihis. 1, Cano de Quiribana. 2, Packeoo Falls. 3, Rupununi. 4, Stondansie, Nickerie
Rivier. 5, headwaters of Coppename. 6, Kaysergebergte. 7, Kleine Saramacca. 8, Suriname Rivier basin (many localities through¬
out the shaded area). 9, Djaikreek. 10, Santarem.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

231

& Axelrod (1968) is so poor and incorrect that only because no other fish
which resembles it closer is known from British Guiana, I accept it as
representing M. poecilus .

Habitat. — A locally numerous inhabitant of medium-sized forest creeks
with running water, in all parts of Suriname. A bottom dweller, mainly found
in places where the bottom is covered with decaying leaves.

Distribution (fig. 41). — Known from Suriname, British Guiana, Vene¬
zuela and Brazil. In Suriname widely distributed (see map and list of mate¬
rial), and known from all major river systems. In British Guiana known from
the Essequibo basin (Packeoo Falls; Rupununi). In Venezuela only known
from Cano de Quiribana, Orinoco. In Brazil there is a single record from
Santarem.

Microglanis variegatus Eigenmann & Henn (fig. 40g)

Microglanis variegatus Eigenmann & Henn, 1914, in Eigenmann, Henn & Wilson,
Indiana Univ. Stud., 19: 14 — forest pool near Vinces, Ecuador.

Microglanis variegatus ; Eigenmann, 1922, Mem. Carnegie Mus,, 9: 33, pi. II fig. 3, 4
(forest pool near Vinces, Ecuador); Gosline, 1941, Stanford Ichth. Bull., 2: 85 (Vinces,
north of Guayaquil, Ecuador) ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33:
28 (oeste do Equador) ; Gomes, 1946, Occ. Pap. Michigan Mus., 494: 8, 9 (forest pool
near Vinces, Provincia Rios, Ecuador) ; Ovchynnyk, 1968, Zool. Anz., 181: 253 (Pool
near Rio Vinces, Prov. Los Rios).

Material. — Five specimens, December 1951, near the mouth of the Rio Cupa, a
tributary of the Rio Blanco, near its junction with the Guaillabanba, Rio Esmeralda basin,
at o°26'N., 79°26'W., Ecuador (Pablo Mena, SU no. 62433), standard length 23, 28, 34,
37, 46 mm.

Characters. — D 1.6 (once I.5), A 11-12, P I.5 or 1 . 6 , V 6 (1.5), C 13-14
(branched rays), branchiostegals 8-9, gill-rakers 6-7 (1 or 2 +1+4 or 5).
A normal representative of the genus, characterized by number of anal rays,
shape of pectoral spine, development of occipital process, shape of caudal fin,
and not in the least place by colour.

Pectoral spine ending in a strong point, along its anterior border a series
of small teeth, along its posterior 'border a series of curved, much larger ones;
as in other members of the genus, the number of teeth increases with growth,
for the smallest specimen has anterior and posterior borders of its spines with
8/6 and 7/6 teeth, the largest with 16/11 and 16/11 teeth; there is no
Y-shaped tooth (fig. 40g).

The band of maxillary teeth is laterally almost square, not so clearly rounded
as in other members of the genus. There is a distinct gap between occipital
process and predorsal plate. The lateral line is developed on the anterior part
of the body only, to above the implantation of the ventrals. The caudal fin is
rounded-truncate (as illustrated by Eigenmann).

23 2

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Colours. The specimens studied were (after 19 years of preservation in
alcohol), more or less even brown, the upper surface of the head and the
predorsal area of the body slightly darker; some specimens show a faint
paler band across the nape, reaching from pectoral to pectoral, and there is
an elongate pale dot before and on the origin of the adipose fin. All fins pale,
lightly pigmented with brown, and without markings except for the caudal
fin which shows somewhat mottled, and has a darker brown base. Unless the
specimens have become strongly decolourized (and they did not give that im¬
pression), by the absence of distinct contrasting markings on body and fins,
this is the dullest member of the genus.

Distribution. — Ecuador west of the Andes, where known from the type-
locality, and from the locality of the sample here listed.

Discussion. — Eigenmann’s (1922: pi. II fig. 4) figure suggests a fish
with a very blunt head and a heavy lower jaw, but this is only seemingly so,
for the specimen has evidently not been photographed in lateral view, but
slightly from below.

Gomes (1946) did not describe this species but in his key gave two charac¬
ters to separate it from all other species: “Lower jaw projecting; head very
wide, broader than long”, against: “Jaws subequal; head longer than broad”.
My material proves that neither of these characters is valid: the lower jaw
of M. variegatus is scarcely protruding, exactly as in its congeners, and the
head is not broader than long, but varies from distinctly longer than broad
to about as broad as long, just as in M. poecilus and M. secundus.

In “broad” specimens of M. parahybae (some of which were examined by
Gomes) and in M. ater, on the other hand, the heads are wider than long, see
descriptions of these species.

Microglanis ater Ahl

Microglanis ater Ahl, 1936, Zool. Anz., 116: 109 — Mittelbrasilien.

Microglanis ater ; Arnold & Ahl, 1936, Fremdl. Siisswasserfische: 248, fig. (Mittel-
und Siidbrasilien); Gosline, 1941, Stanford Ichth. Bull., 2: 85 (Mittelbrasilien) ; Gosline,
1945, Bol. Mus. Nac. Rio de Janeiro, Zool., 33: 29 (Mittelbrasilien) ; Gomes, 1946, Occ.
Pap. Michigan Mus., 494: 9, 16 (Middle Brazil); Fowler, 1951, Arq. Zool. S. Paulo, 6:
530 (Brasil central); Sterba, 1959, Siisswasserfische aus aller Welt: 302, fig. 623 (Mittel-
und Siidbrasilien).

Material. — One specimen, undated, Mittelbrasilien (J. P. Arnold, ZMB no. 20932).
standard length 66 mm, holotype of the species.

Characters. — D 1 . 6 , A 15 (iv.11), P 1 . 6 , V 6 (1.5), C 13 (branched rays
only), branchiostegals 9, gill-rakers 2 + 1 + 7. A large species, easily
distinguished from its congeners by the higher number of anal rays, and dark
colour.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

233

Head 3.6, greatest width of body 3.35, greatest depth of body 4.0, pre¬
dorsal length 2.9 times in standard length. Eyes 3 times in snout, 9 times in
head, ca. 3.5 times in the whole interorbital; maxillary barbels not reaching
to the pectoral base, outer mental barbels of almost the same length, inner
pair about two-thirds of the length of the outer pair, postoccipital process well-
developed, ending in a V-shape, into which the anterior tip of the longer pre¬
dorsal plate is fitted; dorsal spine 2.4 times in predorsal length, pectoral spine
1.6 times in predorsal length.

Colour of preserved specimen. Blackish brown; under surface from chin
to anus white; a faint paler band across the nape; a slightly paler area on
the flanks between the end of the dorsal fin and the anal fin; a pale dot at
the origin of the adipose fin; fins also entirely dark, except for a small white
patch on the last three rays of the dorsal fin, a large, ill-defined white area
covering the basal half of the caudal fin, and a narrow white edge to the
caudal fin. In life the ground-colour would have been black, as it still was
when Ahl (1936) described the specimen.

Distribution. — Supposedly central Brazil, exact localities unknown.

Discussion. — The description made me suspect that this fish belonged in
the relationship of Pseudopimelodus raninus , to which species it shows super¬
ficially a great resemblance, but its dentition shows it to be a true Microglanis .

Microglanis zonatus Eigenmann & Allen

Microglanis zonatus Eigenmann & Allen, 1942, Fish. W. South America: 89, pi. Ill
fig. 1 and 2 — Rio Morona (?).

Microglanis zonatus ; Gosline, 1945, Bol. Mus. Nac. Rio de Janeiro, Z00L, 33: 29
(Amazonas peruano); Fowler, 1945, Los Peces del Peru: 47 (Peru: Rio Morona?);
Gomes, 1946, Occ. Pap. Michigan Mus., 494: 9, 15 (? Rio Morona, Peru); Fowler, 1951,
Arq. Zool. S. Paulo, 6: 531 (Alto Amazonas, Peru).

Material. — None.

Characters. — Very similar to M. poecilus, which it closely resembles in
colour pattern; anterior border of pectoral spines either with all teeth
antrorse, or only the ultimate one retrorse; no Y-shaped tooth; caudal fin
rounded if the published figure represents the specimen correctly.

Distribution. — Eastern Peru, Amazon basin, supposed to be from the
Rio Morona, but this is uncertain.

Discussion. -—- As long as this species remains known from but a single
specimen, of vague provenance, little can be said of its status, except that
description and figure suggest something very close to M. poecilus. Besides
in the rounded caudal fin, the figure of M. zonatus differs in the longer
distance between ventral and anal fin. Possibly M. zonatus is no more than
a subspecies of M . poecilus .

234

ZOOLOGISCHE VERHANDELINGEN 1 3 2 (1974)

Microglanis iheringi Gomes (fig. 40f)

Microglanis iheringi Gomes, 1946, Occ. Pap. Mus. Michigan, 494: 9, pi. I — Rio
Turmero, near Turmero, Aragua, Venezuela.

Microglanis iheringi ; Gray, 1947, Fieldiana, Z00L, 32: 167 (Venezuela: Tumero River
(sic!)) ; Luengo, 1963, Acta Biol. Venez., 3: 321, 331 (Rio Turmero, cerca de Turmero;
Carabobo, Venezuela) ; Mago-Leccia, 1967, Soc. Venez. Cienc. Nat. Bob, 27: 255 (Vene¬
zuela) ; Lundberg & Baskin, 1969, Amer. Mus. Novit., 2398: 6, 28 (no locality).

Material. — One specimen, 13 February 1938, 6 km S.W. of Calabogo, tributary of
the Rio Guarico (F. F. Bond, SU no. 47731), standard length 40 mm.

Characters. — D I.5 1/2, A 12 (v.7), P I.5 1/2, V 6 (i.5), C i.14.1 and
several shorter and rudimentary simple rays, branchiostegals 9, gill-rakers
2+1+4. Similar in general appearance to M. poecilus and M. secundus,
but differs in colour-pattern, structure of the pectoral spine, development
of the lateral line, and larger size.

The pectoral spine is close to that of M. secundus , and different from that
of M. poecilus, in that the antrorse teeth on the posterior edge are large and
well-developed right to the tip; from M. secundus it differs in that teeth along
the anterior edge decrease in size and finally cease distally, and that there
is a more definite tip to the spine, and also probably in the presence of an
Y-shaped tooth on the anterior border of the spine, such as occurs but rarely
in M. secundus . There is presumably some individual variation in structure
of the spine, as indicated by Gomes (1946). Such variation exists also in the
spines of M. poecilus and M. secundus, without affecting their basic differ¬
ences.

The lateral line is well-developed, tubular with pores, to below the adipose
fin; in both M. poecilus and M. secundus the lateral line is well-developed and
tubular to below the dorsal fin only, the posterior part being somewhat rudi¬
mentary, without tubes, although provided with some small pores.

Recorded standard length 26-52 mm (Gomes, 1946). The maximum stan¬
dard length I found in large series of the other two species is 38 mm.

Colours. In colour and colour-pattern generally similar to M. poecilus and
M. secundus, but can be distinguished at a glance from these species by the
mainly dark dorsal fin, which is dark brown, with the tip hyaline, slightly
mottled with brown, and with a large more or less triangular hyaline patch
based on the last ray, and with its apex reaching to the second ray; caudal fin
with a broad dark brown cross J band (M. poecilus has a narrow and often
incomplete crossJband; M. secundus lacks it altogether); the area of dark
brown colour on the caudal peduncle is more or less square, as in M. secun¬
dus, not triangular as in M. poecilus .

Distribution. — Gomes (1946) recorded material from the Rio Turmero,
near Turmero, Aragua, Venezuela, and from the adjacent state of Carabobo,

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

235

Venezuela, without exact locality, and also one specimen from Colombia,
without locality. The last-mentioned specimen is geographically of little use,
except that it indicates a wide distribution. The material examined by Gomes,
as far as I can judge, has been taken from rivers flowing directly north into
the Caribbean, and in the Lago de Valencia. The specimen examined by
me extends the range of the species to the Orinoco basin where, at least
potentially, it co-exists with M. poecilus. This extension of range is not un¬
expected as until about two centuries ago, when its level was 16 m higher than
at present, the Lago de Valencia, of which the Rio Turmero is an affluent,
drained to the Orinoco; for a description and an excellent map of the Lago
de Valencia, see Luengo (1963).

Discussion. — Although the original description is quite comprehensive,
the species has apparently not always been recognised in later years (the
specimen examined by me was received under the name M. poecilus ), and for
that reason I have considered it advisable to give some additional notes.

Microglanis secundus species nova (fig. 40c, pi. 14)

Material. — 104 specimens from Suriname as follows. Three specimens, 28 December
1963, Kwambaolo Kreek, a tributary of the Sara Kreek above Dam (Boeseman, RMNH
no. 26529), standard length 22, 25, 28 mm. One specimen, 6 March 1964, rapids in the
Gran Rio, 1-2 km above its confluence with the Pikien Rio (Boeseman, RMNH
no. 26523), standard length 29 mm. One specimen, 27 November 1965, Tapanahoni, about
2 km upstream of its confluence with the Paloemeu (Mees, RMNH no. 26524), standard
length 2654 mm. One specimen, 23 January 1966, Sipaliwini (Mees, RMNH no. 26525),
standard length 35 mm, holotype of the species. Two specimens, same data (Mees, RMNH
no. 26526), standard length 27, 34 mm. One specimen, 12 October 1966, Sara Kreek,

31 km S. from Dam (Nijssen, ZMA no. 106369), standard length 21 mm, 17 specimens,
14 October 1966, Sara Kreek, 27 km S. from Dam (Nijssen, ZMA no. 105545), standard
length 18-38 mm. 20 specimens, 20 October 1966, Gran Kreek, 63 km S. of Affobakka
(Nijssen, ZMA no. 106002), standard length 1554-34 mm. Eleven specimens, 27 Februrary
1967, tributary of Kleine Saramacca, 11 km E.S.E. from the mouth of the latter (Nijssen,
ZMA no. 106247), standard length 22-35 mm. Five specimens, tributary of Kleine Sara¬
macca, 13 km from the mouth of the latter (Nijssen, ZMA no. 106245), standard length
25, 26, 26, 27, 3654 mm. Six specimens, 20 March 1967, Parwapa Kreek near Botopassie
(Nijssen, ZMA no. 106244), standard length 26, 27, 27, 28, 29, 30 mm. One specimen,
21 March 1967, Jenjee Kreek, 754 km N. of Botopassie (Nijssen, ZMA no. 106243),
standard length 33 mm. Two specimens, 5 April 1967, tributary of Nickerie Rivier, 12 km
W.S.W. from Stondansie (Nijssen, ZMA no. 106242), standard length 27, 2754 mm.

32 specimens, r7 May 1967, Coppename Rivier at 03°49'N., 56°57'W. (Nijssen, ZMA
no. 106246), standard length 23-37 mm. One specimen, same data (Nijssen, ZMA
no. 106211), standard length 30 mm.

Material from British Guiana. One specimen, 3 October 1957, Manari stop-off, Rupu-
nuni District, Amazon drainage (R. H. McConnell, BM no. 1971.7.29: 57), standard
length 21 mm. One specimen, November 1958, Karanambo, Rupununi River, Essequibo
drainage (R. H. McConnell, BM no. 1971.7.29: 52), standard length 2654 mm. One
specimen, 22 April 1961, Wichabai Crossing Pool, Rupununi River (R. H. McConnell,
BM no. 1971.7.29: 53), standard length 22 mm. Two specimens, 24 April 1961, Rupununi

236 ZOOLOGISCHE VERHANDELINGEN I32 (I974)

River and Arakwai Creek (R. H. McConnell, BM no. 19717.29: 50-51), standard length
32J4, 34 mm. Three specimens, 4 May 1961, Semuni tabucas, Rupununi (R. H. McConnell,
BM no. 19717.29: 54-56), standard length 19, 20 y 2 , 21 mm.

Characters. — D I.5 or 1 . 6 , A 8-io, P 1 . 6 , V 6 (i.5), C 12-15 (divided
rays), and several shorter and rudimentary simple rays, branchiostegals 9,
gill-rakers 1 + 1 + 3 or 4. Similar in general appearance to Microglanis
poecilus but differs in colour pattern and in shape and structure of the pec¬
toral spine; size apparently slightly larger; caudal fin of a different shape,
more distinctly forked, with the lobes of roughly equal size; one ray more in
the pectoral fin.

The pectoral spine is quite different from that of M. poecilus. In both
species these spines are somewhat variable, but in M. poecilus the bend in
the spine, where the anterior row of teeth changes direction, is usually marked
by an Y-shaped tooth; moreover towards the tip of the spine, both anterior
and posterior teeth become smaller, and the spine ends in a strong point; in
M. secundus, on the other hand, the bend of the spine where the anterior row
of teeth changes direction is but rarely marked by an Y-shaped tooth (3 out of
94 specimens, and in one of these on one side only), either the change of
direction is abrupt, or there is a simple tooth standing about straight between
the two series; moreover, towards the tip the teeth do not become smaller,
with the result that the tip of the spine does not consist of a point, but of two
teeth, one pointing outwards, one pointing backwards, and the actual tip,
implanted between these teeth, is soft, not spiny. Notwithstanding a certain
amount of individual variation, the differences in shape of the spine,
especially near its tip, are constant in the material examined, they are
illustrated in fig. 40.

Size of the material examined 18-38 mm standard length. Excluding a
single “giant”, the maximum standard length of 756 specimens of M. poecilus
is 33 mm.

Colours. The colour pattern of dark earth brown and light brown is very
similar to that of M. poecilus. The most conspicuous colour differences are:
the pale band across the nape is straighter, the projections forwards and
backwards which give it such an irregular appearance in M. poecilus are
greatly reduced; the saddle on the posterior part of the back is more extensive;
the dark bar across the caudal base is increased in size, covering a square
instead of being triangular, and dorsally it encloses a small but conspicuous
pale spot; there is no dark cross J bar on the caudal fin.

Habitat. — The majority of specimens was taken in the larger creeks, some
in rivers. There, appears to be very little, if any, habitat-difference from M.
poecilus as Nijssen obtained mixed samples of the two species at several

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

237

places (creeks near Botopassie, Kleine Saramacca, Coppename). However,
the general pattern of distribution appears to be more inland than that of M.
poecilus ; the last-mentioned species is plentiful in creeks in the northern
savanna belt, where M. secundus has not been found.

Distribution. — Widely distributed in Suriname, where known from the
basins of the Marowijne (Tapanahoni), Suriname Rivier, Saramacca, Cop¬
pename, Nickerie Rivier, and Corantijn (Sipaliwini). Also Rupununi River,
British Guiana. For details see above under habitat.

Discussion. — Because the two species are likely to be found together, and
to be confused, the preceding description was based on a comparison with M.
poecilus . Of other members of the genus, M. zonatus is also close, but differs
by having a slightly different colour pattern, rounded caudal fin, and having
the teeth on the anterior edge of the pectoral spine all, or practically all,
antrorse. Differences from M. iheringi are given in the description of that
species.

List of Auchenipteridae and Pimelodidae known to occur

in Suriname

Auchenipteridae

Auchenipterus nuchalis (Spix)

Pseudauchenipterus nodosus (Bloch)

Parauchenipterus galeatus (Linnaeus)

Tatia intermedia (Steindachner)

Tatia creutzbergi (Boeseman)

Tatia brunnea n. sp.

Tatia coneolor n. sp.

Tatia punctata n. sp.

Glanidium leopardus (Hoedeman)

Pimelodidae

Hemisorubim platyrhynchos (Valenciennes)

B rachy plat y stoma filamentosum (Lichtenstein)

Brachyplatystoma vaillantii (Valenciennes)

Pseudoplaty stoma fas datum (Linnaeus)

Pimelodus ornatus Kner
Pimelodus blochii Valenciennes
Pimelodus albofasciatus n. sp.

Pimelodella cristata (Muller & Troschel)

238

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Pimelodella macturki Eigenmann
Rhamdia quelen (Quoy & Gaimard)

Imparfinis minutus (Liitken)

Heptapterus brevior (Eigenmann)

Heptapterus longior (Eigenmann)

Heptapterus tapanahoniensis Mees
Heptapterus surinamensis Bleeker
Heptapterus bleekeri Boeseman
Pseudopimelodus raninus raninus (Valenciennes)
Pseudopimelodus raninus villosus Eigenmann
Pseudo pimelodus zungaro bufonius (Valenciennes)
Pseudo pimelodus nigricauda n. sp.

Microglanis poecilus Eigenmann
Microglanis secundus n. sp.

New synonymy

Auchenipterus ambyiacus Fowler,
1915

Auchenipterus brevior Eigenmann,
1912

Auchenipterus demerarae
Eigenmann, 1912

Auchenipterus glaber Steindachner,
1876

Auchenipterus obscurus Gunther,
1863

Brachyplatystoma goeldii
Eigenmann & Bean, 1907

Caecorhamdia Norman, 1926

Centromochlus dunni Fowler, 1945
(possibly a valid race)

Centromochlus steindachneri Gill,
1871

Cephalosilurus Haseman, 1911

Ceratocheilus osteomystax
A. de Miranda Ribeiro, 1918

Cetopsorhamdia pijpersi Hoedeman,
1961

Gephyromochlus Hoedeman, 1961

= Auchenipterus nuchalis

})

= Par auchenipterus galeatus

= Trachycorystes trachycorystes

= Brachyplatystoma filamentosum

= Rhamdia
= Tatia intermedia

= Centromochlus heckelii

= Pseudo pimelodus
= Auchenipterus nuchalis

— Imparfinis minutus

= Glanidium

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

239

Glanidium cesarpintoi
R. v. Ihering, 1928
Imparales Schultz, 1944
Imparfinis hasemani Steindachner,
1915

Medemichthys Dahl, 1961
Nannorhamdia Regan, 1913
Pariolius Cope, 1872
Phenacorhamdia Dahl, 1961
Pimelodella geryi Hoedeman, 1961
Pimelodella insignis Schubart, 1964
Pimelodus altipinnis Steindachner,
1864

Pimelodus char us Valenciennes, 1840

Pimelodus eigenmanni
v. d. Stigchel, 1946
Pimelodus pulcher Boulenger, 1887
Pseudo pimelodus roosevelti
Borodin, 1927
Pseudo pimelodus variolosus
A. de Miranda Ribeiro, 1914
Pseudo pimelodus villosus butcheri
Schultz, 1944

Rhamdia holomelas rupununi
Fowler, 1914

Trachycorystes coracoideus
Eigenmann & Allen, 1942
Trachycorystes jokeannae
Hoedeman, 1961
Zungaro Bleeker, 1858

= f Glanidium ribeiroi

— Heptapterus

— Imparfinis minutus

= Heptapterus

— Imparfinis
Heptapterus

— Heptapterus

= Pimelodella cristata
= Pimelodella boschmai
“ Pimelodella cristata

= Pseudo pimelodus zungaro
bufonius

— Pimelodus altissimus

— Pseudopimelodus z. zungaro

— Pseudopimelodus zungaro

mangurus

— Pseudopimelodus z. zungaro

= Pseudopimelodus raninus
acanthochiroides

— Rhamdia foina

— Auchenipterichthys thoracatus
= Parauchenipterus galeatus

— Pseudopimelodus

Samenvatting

Tot voor nauwelijks tien jaren was de vis fauna van Suriname nog zeer
onvoldoende bekend, doch in de jaren 1963/1967 werden grote verzamelingen
bijeengebracht, achtereenvolgens door M. Boeseman, G. F. Mees, en H. Nijs-
sen. Deze collecties bevinden zich in het Rijksmuseum van Natuurlijke
Historic te Leiden en het Zoologisch Museum te Amsterdam; zij zijn af-
komstig uit de stroomgebieden van de meeste belangrijke rivieren, en al zullen
zij zeker nog geen volledig beeld van de visfauna geven, toch valt te ver-

240

ZOOLOGISCHE VERHANDELINGEN I32 (I974)

wachten dat als deze verzamelingen volledig bewerkt zijn (iets dat overigens
nog jaren zal kosten), Suriname een der ichthyologisch het best bekende
landen van Zuid-Amerika zal zijn geworden.

In dit artikel wordt een overzicht gegeven van de van Suriname bekende
vertegenwoordigers der Auchenipteridae en Pimelodidae, twee families be-
horende tot de in Zuid-Amerika zeer soortenrijke orde der meervalien (Nema-
tognathi of Silurif ormes).

Tijdens het onderzoek bleek al spoedig dat een serieuze studie van de Suri-
naamse soorten niet mogelijk was zonder er literatuur en materiaal uit om-
ringende landen bij te betrekken. In verschillende gevallen resulteerde dit in
nogal ingrijpende wijzigingen van de gangbare soort- en geslachtsomgren-
zingen en daarmede samenhangend van de classificatie. In dit artikel worden,
uit verschillende delen van Zuid-Amerika, 3 nieuwe genera en 14 nieuwe
soorten beschreven, terwijl 2 vroeger beschreven genera die in recente litera¬
tuur niet meer werden erkend, uit de synonymie zijn genomen en nieuw
worden gedefinieerd. Daar tegenover staat dat q eerder beschreven genera
en 24 soorten voor het eerst in de synonymie moesten worden geplaatst.

Daar, in de uitgebreide behandeling van bepaalde genera, vaak de gegevens
over Suriname dreigden te verdrinken, is terwille van de overzichtelijkheid
een lijst toegevoegd van de thans van Suriname bekende soorten (biz. 237-238).
De Auchenipteridae worden vertegenwoordigd door 9 soorten, behorende
tot $ genera, terwijl van de Pimelodidae 21 soorten behorende tot 10 genera
bekend zijn; een dezer soorten wordt vertegenwoordigd door twee onder-
scheidbare geographische rassen.

References

Agassiz, L. & Mme Agassiz, 1869. Voyage au Bresil (traduit de 1 ’anglais avec l'auto-
risation des auteurs). Paris: i-xii, 1-532.

Ahl, E., 1936. Beschreibungen dreier neuer Welse aus Brasilien. — Zool. Anz., 116:
109-111.

Aramburu, A. A. de, R. H. Aramburu & R. A. Ringuelet, 1962. Peces paranenses
nuevos para la fauna Argentina. — Physis, 23: 222-239.

Arnold, J. P. & E. Ahl, 1936. Fremdlandische Siisswasserfische. Braunschweig: r-592.
Axelrod, H. R. & L. P. Schultz, 1955. Handbook of tropical aquarium fishes. New
York, Toronto, London: i-xii, 1-718.

Beebe, W., 1925. Studies of a tropical jungle; one quarter of a square mile of jungle at
Kartabo, British Guiana. — Zoologica, 6: 5-193.

Berg, C., 1901. Communicaciones ictiologicas, IV. — Comm. Mus. Nac. Buenos Aires, 1:
293-311.

Bertin, L. & R. Esteve, 1950. Catalogue des types de poissons du Museum National
d’Histoire Naturelle. 5 e Partie. Ostariophysaires (Siluriformes). Paris: 1-85.
Bleeker, P., i 862(- i 863). Atlas Ichthyologique, 2. Amsterdam: 1-112.

-, 1863. Systema silurorum revisum. — Ned. Tijdschr. Dierk., 1: 77-122.

-, 1864. Description des especes de Silures de Suriname conservees aux musees de

Leide et d*Amsterdam. — Nat. Verh. Holl. Maatsch. Wetensch., (2) 20: 1-104.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

241

Bloch, M. E., 1785. Ichthyologie ou histoire naturelle generate et particuliere des pois-
sons, 1. Berlin: 1-206.

Boeseman, M., 1952. A preliminary list of Surinam fishes not included in Eigenmann's
enumeration of 1912. — Zool. Meded., 31: 179-200.

-» 1953 - Scientific results of the Surinam expedition 1948-1949. Part II. Zoology No. 2.

The fishes (1). — Zool. Meded., 32: 1-24.

-, i960. The fresh-water fishes of the island of Trinidad. — Studies on the Fauna

of Curasao and other Caribbean Islands, 48: 72-153.

-, 1968. The genus Hypostomus Lacepede, 1803, and its Surinam representatives

(Siluriformes, Loricariidae). — Zool. Verh., 99: 1-89.

-, 1969. Additional new species of Hypostomus Lacepede, 1803, from Surinam; with

remarks on the apparent “gymnorhynchus-complex” (Siluriformes, Loricariidae). —
Beaufortia, 16: 119-136.

-, 1972. Notes on South American catfishes, including remarks on Valenciennes and

Bleeker types in the Leiden museum. — Zool. Meded., 47: 293-320.

Borodin, N. A., 1927. Some new catfishes from Brazil. — Amer. Mus. Novit., 266: 1-7.

Boulenger, G. A., 1887. An account of the fishes collected by Mr. C. Buckley in eastern
Ecuador. — Proc. Zool. Soc. Lond.: 274-283, pi. XX-XXIV.

-, 1891. An account of the siluroid fishes obtained by Dr. H. von Ihering and Herr

Sebastian Wolff in the province Rio Grande do Sul, Brazil. — Proc. Zool. Soc.
Lond.: 231-235.

Britski, H. A., 1969. Lista dos tipos de peixes des cole^oes do Departamento de Zoologia
da Secretaria da Agricultura de Sao Paulo. — Pap. Avuls. Zool. S. Paulo, 22:
197-215.

Bussing, W. A., 1970. Two new species of catfishes of the genera Nannorhamdia and
Imparales (family Pimelodidae) from Central America. — Contrib. Sci. Los Angeles
County Mus., 196: 1-11.

Carriker, M. A., 1931. The cave birds of Trinidad. — Auk, 48: 186-194.

Castello, H. P., 1971. Pseudopimelodus raninus (Valenciennes 1840) una nueva cita
para la fauna de peces de la Republica Argentina. — Physis, 30: 543 - 545 -

Cope, E. D., 1878. Synopsis of the fishes of the Peruvian Amazon obtained by Professor
Orton during his expeditions of 1873 and 1877. — Proc. Amer. Philos. Soc., 17:
673-701.

Cuvier, G. & A. Valenciennes, 1840. Histoire naturelle des poissons (4 0 ed.), 15.
Paris: i-xxiv, 1-397.

Dahl, G., 1961. Nematognathous fishes collected during the Macarena expedition 1959.
Part II — Noved. Colomb., 1: 483-514.

-, 1971. Los Peces del Norte de Colombia. Bogota: i-xvii, 1-391.

Devincenzi, G. J., 1933. Peces del Uruguay. Notas complementarias, II. — An. Mus.
Hist Nat. Montevideo, (2) 4 (3): 1-11, 1 pi.

Devincenzi, G. J. & L. P. Barattini, 1926-1940. Album ictiologico del Uruguay. —
An. Mus. Hist Nat. Montevideo, Supl.: 1-8, pis. I-LII.

Devincenzi, G. J. & G. W. Teague, 1942. Ictiofauna del Rio Uruguay Medio. — An.
Mus. Hist. Nat. Montevideo, (2) 5: I-VIII, 1-104.

Eigenmann, C. H., 1911. The localities at which Mr. John D. Haseman made collec¬
tions. — Ann. Carnegie Mus., 7: 299-314.

-, 1912. The freshwater fishes of British Guiana, including a study of the ecological

grouping of species and the relation of the fauna of the plateau to that of the
lowlands. — Mem. Carnegie Mus., 5: i-xvii, 1-578.

-, 1917. Pimelodella and Typhlobagrus. — Mem. Carnegie Mus., 7: 229-258.

-, 1922. The fishes of western South America, part. I. — The fresh-water fishes

of northwestern South America including Colombia, Panama, and the Pacific slopes
of Ecuador and Peru, together with an appendix upon the fishes of the Rio Meta in
Colombia. — Mem. Carnegie Mus., 9: 1-346.

16

242 ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Eigenmann, C. H. & W. R. Allen, 1942. Fishes of western South America. Kentucky:
i-xv, 1-494.

Eigenmann, C. H. & B. A. Bean, 1907. An account of Amazon River fishes collected
by J. B. Steere; with a note on Pimelodus clarias. — Proc. U.S. Nat. Mus., 31:
659-668.

Eigenmann, C. H. & R. S. Eigenmann, 1888. Preliminary notes on South American
Nematognathi I. — Proc. Calif. Acad. Sri., (2) 1: 119-172.

-, 1889. Preliminary notes on South American Nematognathi II. — Proc. Calif.

Acad. Sci., (2) 2: 28-56.

-, 1890. A revision of the South American Nematognathi or cat-fishes. — Occ. Pap.

Calif. Acad. Sci., 1: 1-508.

Eigenmann, C. H. & A. A. Norris, 1900. Sobre alguns peixes de S. Paulo, Brazil. —
Rev. Mus. Paul., 4: 349-362.

Emmens, C. W. & H. A. Axelrod, 1968. Catfish. New York: 1-96.

Fernandez Yepez, A. & F. Martin S., 1953. Apuntes sobre la ictiologia de Perija. —
La region de Perija y sus habitantes (Soc. Cienc. Nat. La Salle, 13) : 299-313.

Filippi, Ph. de, 1853. Nouvelles especes de poissons. — Rev. Mag. Zool., (2) 5: 164-171.

Fisher, H. G., 1917. A list of the Hypophthalmidae, the Diplomystidae, and of some
unrecorded species of Siluridae in the collections of the Carnegie Museum. — Ann.
Carnegie Mus., 11: 405-427.

Fowler, H. W., 1914. Fishes from the Rupununi River, British Guiana. — Proc. Acad.
Nat. Sci. Philad., 66: 229-284.

-, 1915. Notes on nematognathous fishes. — Proc. Acad. Nat. Sci. Philad., 67: 203-243.

-, 1940a. A collection of fishes obtained by Mr. William C. Morrow in the Ucayali

River basin, Peru. — Proc. Acad. Nat. Sci. Philad., 91: 219-289.

-, 1940b. Zoological results of the second Bolivian expedition for the Academy of

Natural Sciences of Philadelphia, 1936-1937. Part I. — The Fishes. — Proc. Acad.
Nat. Sci. Philad., 92: 43-103.

-, 1941a. A collection of fresh-water fishes obtained in eastern Brazil by Dr. Rodolpho

von Ihering. — Proc. Acad. Nat. Sci. Philad., 93: 123-199.

-, 1941b. Los peces del Peru (Continuacion). — Bol. Mus. “Javier Prado”, 5: 466-487.

-, 1945a. Colombian zoological survey. Part I. — The fresh-water fishes obtained in

1945. — Proc. Acad. Nat. Sci. Philad., 97: 93-135.

-, 1945b. Los Peces del Peru. Lima: 1-298.

-, 1951. Os peixes de agua doce do Brasil, 1 (3). — Arq. Zool. S. Paulo, 6: 405-625.

Geijskes, D. C., 1973. Reisverslag van de expeditie West Suriname 1971. — Zool. Bijdr.,
15: i* 4 i.

Gill, Th., 1871. On some new species of fishes obtained by Prof. Orton from the Mara-
non, or Upper Amazon, and Napo Rivers. — Proc. Acad. Nat. Sci. Philad., 1870:
92-96.

Goeldi, E. A., 1901. A Piralba, gigantesco siluroideo do Amazonas. — Bol. Mus. Para-
ense, 3: 181-194.

Gomes, A. L., 1946. A review of Microglanis, a genus of South American catfishes,
with notes on related genera. — Occ. Pap. Mus. Zool. Michigan, 494: 1-19.

- , 1956. Descri^ao de uma nova esperie de “Luciopimelodinae” do Rio Mogi Gua$u,

Estado de Sao Paulo. — Rev. Brasil. Biol., 16: 403-413.

Gosline, W. A., 1940. Rediscovery and redescription of Pariolius armillatus, a genus
and species of pimelodid catfishes described by E. D. Cope from the Peruvian
Amazon in 1872. — Copeia: 78-80.

—, 1941. Synopsis of the genera of pimelodid catfishes without a free orbital rim. —
Stanford Ichth. Bull., 2: 83-88.

-, 1942. Notes on South American catfishes (Nematognathi). — Copeia: 39-41.

-, 1945. Catalogo dos nematognathos de agua-doce da America do Sul e Central. —

Bol. Mus. Nac. (Rio de Janeiro), Zool., 33: 1-138.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

243

Guntert, H, 1942. Beschreibung einiger zum Teil noch unbekannter siidamerikanischer
Siluriden aus dem Naturhistorischen Museum in Basel. — Zool. Anz., 138: 27-40.
Gunther, A., 1863. On new species of fishes from the Essequibo. — Ann. Mag. Nat.
Hist, (3) 12: 441-443.

-, 1864. Catalogue of the fishes in the British Museum, 5. London: i-xxii, 1-455.

Haseman, J. D, 1911. Descriptions of some new species of fishes and miscellaneous
notes on others obtained during the expedition of the Carnegie Museum to central
South America. — Ann. Carnegie Mus, 7: 315-328.

Heide, J. van der, 1966. Het stuwmeeronderzoek in Suriname. — Tijdschr. Kon. Ned.
Aardr. Gen, (2) 83: 173-180.

Hoedeman, J. J, 1954. Pimelodus clarias. — Het Aquarium, 24: 172-173.

-, 1957. Notes on the ichthyology of Surinam (Dutch Guiana), 4. Additional records

of siluriform fishes (1). — Beaufortia, 6: 147-160.

-, 1961. Notes on the ichthyology of Surinam and other Guianas. — Bull. Aquatic

Biol, 2: 129-139.

-, 1968. Elseviers Aquariumvissen Encyclopedic, 1. Amsterdam/Brussel: 1-188.

Holthuis, L. B, 1959. The Crustacea Decapoda of Suriname. — Zool. Verh, 44: 1-296.
Hubbs, C. L, 1936. Fishes of the Yucatan Peninsula. — Carnegie Mus. Wash. Publ,
457 : 157-287.

-, 1938. Fishes from the caves of Yucatan. — Carnegie Inst. Wash. Publ, 491: 261-

295, pi. 1-4.

Ihering, R. von, 1930. Notas ecologicas referentes a peixes d’ague doce do Estado de
S. Paulo e descrip^ao de 4 especies novas. — Arch. Inst. Biol. S. Paulo, 3: 93-104,
pi. 13.

-, 1937. Oviducal fertilization in the South American catfish, Trachycorystes. —

Copeia: 201-205.

Jentink, F. A, 1910. Verslag omtrent ’$ Rijks Museum van Natuurlijke Historic te
Leiden, loopende over het tijdvak van 1 September 1909 tot 1 September 1910.
Leiden: 1-27.

Kappler, A, 1881. Hollandisch-Guiana, Stuttgart: I-VIII, 1 - 495 -

-, 1885. Die Tierwelt im Hollandischen Guiana (Fortsetzung). — Das Ausland, 58:

896-899, 918-920.

-, 1887. Surinam, sein Land, seine Natur, Bevolkerung und seine Kultur-Verhaltnisse.

Stuttgart: 1-384.

Kner, R, 1858. Ichthyologische Beitrage. — Sitzb. Akad. Wiss. Wien, Mathem.-Naturw.
Cl, 26 (1857) : 373-448.

La Cep£de, 1803. Histoire naturelle des poissons, 5. Paris: i-lxviii, 1-803.
Lichtenstein, M, H. C, 1819. Ueber einige neue Arten von Fischen aus der Gattung
Silurus. — Zool. Mag. (Wiedemann), 1 (3): 57-63.

Linnaeus, C, 1766. Systema naturae, ed. 12, 1. Holmiae: 1-532.

Lowe (McConnell), R. H, 1964. The fishes of the Rupununi savanna district of
British Guiana, South America. Part 1. Ecological groupings of fish species and
effects of the seasonal cycle on fish. — J. Linn. Soc, Zool, 45: 103-144.

-•, 1967. The fish-fauna of the Rupununi District, Guyana — Timehri, 43: 58-72.

Luengo, J. A, 1963. La fauna ictiologica del Lago de Valencia (Venezuela) y algunas
consideraciones sobre las demas hoyas del pais y Trinidad. — Acta Biol. Venez, 3:
319 - 339 -

Luling, K. H, 1963. Die Quisto Cocha und ihre haufigen Fische. — Beitr. Neotrop.
Fauna, 3: 34-56.

Lundberg, J. G. & J. N. Baskin, 1969. The caudal skeleton of the catfishes, order
Siluriformes. — Amer. Mus. Novit, 2398: 1-49.

Lutken, Chr. Fr, 1875. Velhas-Flodens fiske. — Vidensk. Selsk. Skr, (5) 12: 124-254,
I-XXI, pi. I-V.

244

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Lynden, A. J. H. van, 1939. Op zoek naar Suriname’s zuidgrens. — Tijdschr. Kan. Ned.
Aardr. Gen., (2) 56: 793-871, phot. 1-91, maps XVI-XVIII.

McConnell, R. H., see under Lowe.

Mago-Leccia, F., 1967. Notas preliminares sobre los peces de los llanos de Venezuela. —
Bol. Soc. Venez. Cienc. Nat., 27: 237-263.

Marlier, G., 1968. Etudes sur les lacs de l’Amazonie centrale. II. Le plancton. III. Les
poissons du lac Redondo et leur regime alimentaire; les chaines tropiques du lac
Redondo; les poissons du Rio Preto da Eva. — Cad. Amazonia, 11: 1-57.

Mees, G. F., 1957. A systematic review of the Indo-Australian Zosteropidae (Part I).
Leiden: 1-204, Stellingen I-XI.

-, 1962. The subterranean freshwater fauna of Yardie Creek Station, North West

Cape, Western Australia. — J. Roy. Soc. W. Austr., 45: 24-32.

-, 1967. Freshwater fishes of Suriname: the genus Heptapterus (Pimelodidae). —

Zool. Meded., 42: 215-229.

Miles, C, 1947. Los peces del Rio Magdalena. Ministerio Econ. Nac., Bogota: 1-214,
i-xxviii.

Myers, G. S., 1928. New fresh-water fishes from Peru, Venezuela, and Brazil. — Ann.
Mag. Nat. Hist., (10) 2: 83-90.

-, 1947. The Amazon and its fishes ... Part 2. The fishes. — The Aquarium J., 18

(4): 13-20.

Norman, J. R., 1926. A new blind catfish from Trinidad, with a list of the blind cave-
fishes. — Ann. Mag. Nat. Hist, (9) 18: 324-331.

Nijssen, H., 1970. Revision of the Surinam cat fishes of the genus Corydoras Lacepede,
1803. Amsterdam: 1-75, I-II.

Nijssen, H. & I. J. H. Isbrucker, 1967. Notes on the Guiana species of Corydoras
Lacepede, 1803, with descriptions of seven new species and designation of a neotype
for Corydoras punctatus (Bloch, 1794) — (Pisces, Cypriniformes, Callichthyidae). —
Zool. Meded., 42: 21-50.

Pavan, C., 1946. Observations and experiments on the cave fish Pimelodella kronei and
its relatives. — Amer. Nat., 80: 343-361.

Pearson, N. E., 1924. The fishes of the eastern slope of the Andes. I. The fishes of
the Rio Beni Basin, Bolivia, collected by the Mulford expedition. Indiana Univ.
Stud., 11 (64) : 1-83.

Pellegrin, J., 1899. Notes sur les poissons recueillis par M. F. Geay dans l’Apure et ses
affluents. — Bull. Mus. Hist. Nat. Paris, 5: 156-159.

Pelzeln, A. von, 1868-1871. Zur Omithologie Brasiliens. Wien: 1-462, I-LIX, 1-18.

Peters, W., 1877. t)ber die von Hra. Dr. C. Sachs in Venezuela gesammelten Fische. —
Mber. K. Akad. Wiss. Berlin: 469-473.

Puyo, J., 1949. Poissons de la Guyane Frangaise. Faune de l’Empire Frangais XII:
1-280.

Regan, C. T., 1906. On the fresh-water fishes of the island of Trinidad, based on the
collection, notes, and sketches made by Mr. Lechmere Guppy, Junr. — Proc. Zool.
Soc. Lond: 378-393.

-, 1911. The classification of the teleostean fishes of the order Ostariophysi. — 2. Silu-

roidea. — Ann. Mag. Nat. Hist., (8) 8: 553 - 577 -

-, 1913. The fishes of the San Juan River, Colombia. — Ann. Mag. Nat. Hist., (8)

12: 462-473.

Ribeiro, A. de Miranda, 1911. Fauna Brasiliense. Peixes IV (A): Eleutherobranchios
Aspirophoros. — Arch. Mus. Nac. Rio de Janeiro, 16: 1-504.

-1914. Pimelodidae, Trachycorystidae, Cetopsidae, Bunocephalidae, Auchenipteridae,

Hypophthalmidae. — Comm. Linh. Telegr. Estr. de Matto-Grosso ao Amazonas, 15.
Annexo 5. Historia Natural. Zoologia: 1-13, pi. 1, 2.

-, 1918a. Tres generos e dezesete especies novas de peixes brasileiros. — Rev. Mus.

Paul., 10: 629-646.

MEES, AUCHENIPTERIDAE AND PIMELODIDAE

245

Ribeiro, A. be Miranda, 1918b. Lista dos peixes brasileiros do Museu Paulista (i.a
Parte). — ReW Mus. Paul., 10: 705-736.

-, 1920. Peixes (excl. Characinidae). — Comm. Linh, Telegr. Estr. de Matto-Grosso

ao Amazonas, 58. Annexo 5. Historia Natural. Zoologia: 1-15, 17 pis.

—, 1937- Sobre una collecgao de vertebrados do nordeste brasileiro. Primeira parte:
Peixes e batrachios. — O Campo, 1: 54-56.

Ribeiro, P. de Miranda, 1953. Tipos das especies e subespecies do Prof. Alipio de Mi¬
randa Ribeiro depositados no Museu Nacional. —- Arq. Mus. Nac. Rio de Janeiro,
42: 389-417.

-, 1962a. Apontamentos ictiologicos, I. — Bol. Mus. Nac. Rio de Janeiro, Zool.,

240: 1-6.

-, 1962b. Catalogo dos peixes do Museu Nacional, XI. — Publ. Avuls. Mus. Nac.

Rio de Janeiro, 45: 1-12.

—, 1964a. Apontamentos ictiologicos, II. — Bol. Mus. Nac. Rio de Janeiro, Zool.,
246: 1-4.

-, 1964b. Rhamdella schultzi sp. nov. — Bol. Mus. Nac. Rio de Janeiro, 248: 1-4.

-, 1968a. Sobre o dimorfismo sexual no genero Auchenipterus Valenciennes, 1840. —

Bol. Mus. Nac. Rio de Janeiro, Zool., 261: 1-11.

-, 1968b. Apontamentos ictiologicos, III. — Bol. Mus. Nac. Rio de Janeiro, Zool.,

263: 1-14.

-, 1968c. Apontamentos ictiologicos, V. — Bol. Mus. Nac. Rio de Janeiro, 264: 1-5.

Ridgway, R., 1912. Color standards and color nomenclature. Washington, D.C.: i-iv,

1-44, pl- I-LIII.

Ringuelet, R. A., R. H. ArAmburu & A. A. de ArAmburu, 1967. Los peces Argentinos
de agua dulce. La Plata: 1-602.

Rossel, F., 1962. Centromochlus schultzi, ein neuer Weis aus Brasilien. — Senck. Biol.,
43 : 27-30.

Schomburgk, R., 1848. Reisen in Britisch-Guiana in den Jahren 1840-1844, 3. Leipzig:
i-viii, 533-1261.

Schomburgk, R. H., 1841. Fishes of Guiana, I. The Naturalist’s Library, Ichthyology,
3: 1-262, pi. I-XXX.

Schubart, O., 1964. Duas novas especies de peixe de familia Pimelodidae do Rio Mogi
Guagu (Pisces, Nematognathi). — Bol. Mus. Nac. Rio de Janeiro, Zool., 244: 1-22.

Schultz, L. P., 1944a. The catfishes of Venezuela, with descriptions of thirty-eight new
forms. — Proc. U.S. Nat. Mus., 94: 173-338.

-, 1944b. A new genus and species of pimelodid catfish from Colombia. — J. Wash.

Acad. Sci., 34: 93-95.

Seba, A., 1758. Nauwkeurige beschryving van het schatryke kabinet der voornaamste
seldzaamheden der natuur (Locupletissimi rerum naturalium thesauri), 3. Amstelae-
dami: 1-212.

Sherborn, C. D. & B. B. Woodward, 1901. Notes on the dates of publication of the
natural history portions of some French voyages, I. “Amerique meridionale”; “Indes
orientates”; “Pole Sud” (“Astrolabe” and “Zelee”); “La Bonite”; “La Co-
quille”; and “L’Uranie et Physicienne”. — Ann. Mag. Nat. Hist. (7) 7: 388-392.

Spix, J. B. de, 1829. Selecta genera et species piscium. Monachii. I-II, i-xvi, I-II,
1-138, Tab. A-G, I- LXXV, Anat. A-F.

Starks, E. C., 1913. The fishes of the Stanford expedition to Brazil. — Stanford Publ.
Univ. Ser.: 1-77, pl. I-XV.

Steenis-Kruseman, M. J. van, 1950. Flora Malesiana, (1) 1: I-CLII, 1-639.

Steindachner, F. f 1864. Ichthyologische Notizen. — Sitzb. Akad. Wiss. Wien, Mathem.-
Naturw. Cl., 49 (1): 200-214, Taf. 1, 2.

-, 1876a. Ichthyologische Beitrage (IV). — Sitzb. Akad. Wiss. Wien, Mathem.-

Naturw. Cl., 72 (1) : 551-616, pl. I-XIII.

246

ZOOLOGISCHE VERHANDELINGEN I32 (1974)

Steindachner, F., 1876b. Die Siisswasserfische des sudostlichen Brasilien (III). — Sitzb.
Akad. Wiss. Wien, Mathem.-Naturw. CL, 74 (1): 559-694

-, 1879. Beitrage zur Kenntnis der Flussfische Siidamerika’s. — Denkschr. Akad.

Wiss. Wien, Mathem.-Naturw. CL, 41 (1) : 151-172.

—, 1880. Zur Fisch-Fauna des Cauca und der Fliisse bei Guayaquil. — Denkschr.

Akad. Wiss. Wien, Mathem.-Naturw. CL, 42 (1) : 55-104.

—, 1882. Beitrage zur Kenntnis der Flussfische Siidamerika’s, III. — Denkschr. Akad.
Wiss. Wien, Mathem.-Naturw. CL, 44 (x): 1-18.

-, 1883. Beitrage zur Kenntnis der Flussfische Sudamerika’s, IV. — Denkschr. Akad.

Wiss. Wien, Mathem.-Naturw. CL, 46 (1): 1-44.

-, 1915. Beitrage zur Kenntnis der Flussfische Sudamerika’s. — Denkschr. Akad.

Wiss. Wien, Mathem.-Naturw. KL, 93: 15-106, pi. I-XIII.

Stigchel, J. W. B. van der, 1946. South American Nematognathi. Leiden: 1-204 (also,
with identical pagination, in Zool. Meded., 27 (1947)).

-, 1964. A new species of pimelodid catfish from eastern Brazil, Pimelodella boschmai

nov. spec — Zool. Meded.. 39: 327-330.

Things, G., 1955. Les poissons aveugles (I). — Ann. Soc. Roy. Zool. Belg., 86: 1-128.

-, 1969. L’evolution regressive der poissons cavemicoles et abyssaux. Paris: 1-394.

Vaillant, L., 1880. Synopsis des especes de Siluridae recueillies par M. le Dr Jobert,
a Calderon (Haute-Amazone). — Bull. Soc. Philomat. Paris, (7) 4: 150-159.

-, 1899. Note preliminaire sur les collections ichthyologiques recueillies par M. Geay

en 1897 et 1898 dans la Guyane frangaise et le Conteste franco-bresilien. — Bull.
Mus. Hist. Nat. Paris, 5: 154-156.

-, 1900. Contribution a l'etude de la faune ichthyologique de la Guyane Frangaise et

du Conteste franco-bresilien. — Nouv. Arch. Mus. Hist. Nat. Paris, (4) 2: 123-136,

pL 7 -

Valenciennes, A., 1840. See under Cuvier & Valenciennes.

-, 1847. Poissons. In A. d’Orbigny: Voyage dans l’Amerique Meridionale, 5 (2) :

1-11; 9 (Atlas), Poiss.: pi. 1-16 (plates published in 1834-1837).

Westermann, J. H., 1971. Historisch overzicht van de wording en het onderzoek van
het Brokopondo-stuwmeer. — Nieuwe West-Indische Gids, 48: 1-55.

ZOOLOGlSCHE VERHANDELINGEN 132 (1974)

Pl. 1

[V

Trachycorystes trachycorystes (MNP no. A 9422, s.l. ca. 310 mm).

ZOOLOGTSCHE \ ’ EE HAN D EL IN GEN 132 (1974)

Pl. 2

c -a

jy ^

&

O ^ 'o
P PC 1
. . C -O
:p ^ p

t S s

Ph
<y rt

O *S
C

->

S tj- iu
>->

U o 2

.. P "

Mh . *2

H £ -3

K r-» ^

£ <L> ><

O Ph £

a

.3 ^
&> ..
+-I

be

s 5

1 .e

. t-H

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Pl. 3

ZOOLOGISCHE VF.RHANDELINGEN 132 (1974)

Pi- 4

Pimelodus albofasciatus (RMNH no. 26516). 1 X.

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Pl. 5

Pscndopimclodus zungaro. Reproduction of Humboldt’s figure, at the same size as ii

the original publication.

ZOOLOGISCHF. VF.RHANDFLINGFN 132 (1974)

Pl. 6

ZOOLOGTSCHE VERHANDELINGEN 132 (1974)

Pl. 7

slender body and the anteriorly rounded head; the teeth on the pectoral spines are

clearly visible.

ZOOLOGISCHE YERHANDELINGEN 132 (1974)

Pl. 8

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Pl. 9

Pseudopimelodus raninus raninus (same specimen as in plate 8 ) ; note, compared with
P. zungaro, the broader, heavier head and body.

ZOOLOGISCHE VEKHANDELINGEN 132 (1974) Pl. 12

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Pl. 13

ZOOLOGISCHE VERHANDF.LINGEN 132 (1974) Pl. 14

ZOOLOGISCHE VERHANDELINGEN 132 (1974)

Pi- 15

Brachyplatystoma vaillantii , specimen of ca. 170 cm total length (Brokopondo, 26-II-1964,

phot. A. Bakker).